CA1326520C - Locomotive wheelslip control system - Google Patents

Locomotive wheelslip control system

Info

Publication number
CA1326520C
CA1326520C CA000607117A CA607117A CA1326520C CA 1326520 C CA1326520 C CA 1326520C CA 000607117 A CA000607117 A CA 000607117A CA 607117 A CA607117 A CA 607117A CA 1326520 C CA1326520 C CA 1326520C
Authority
CA
Canada
Prior art keywords
value
speed
wheelslip
predetermined
operative
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
CA000607117A
Other languages
French (fr)
Inventor
Edgar Thomas Balch
Harold Stevenson Hostettler, Jr.
David John Konko
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
General Electric Co
Original Assignee
General Electric Co
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by General Electric Co filed Critical General Electric Co
Application granted granted Critical
Publication of CA1326520C publication Critical patent/CA1326520C/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B60VEHICLES IN GENERAL
    • B60TVEHICLE BRAKE CONTROL SYSTEMS OR PARTS THEREOF; BRAKE CONTROL SYSTEMS OR PARTS THEREOF, IN GENERAL; ARRANGEMENT OF BRAKING ELEMENTS ON VEHICLES IN GENERAL; PORTABLE DEVICES FOR PREVENTING UNWANTED MOVEMENT OF VEHICLES; VEHICLE MODIFICATIONS TO FACILITATE COOLING OF BRAKES
    • B60T8/00Arrangements for adjusting wheel-braking force to meet varying vehicular or ground-surface conditions, e.g. limiting or varying distribution of braking force
    • B60T8/17Using electrical or electronic regulation means to control braking
    • B60T8/1701Braking or traction control means specially adapted for particular types of vehicles
    • B60T8/1705Braking or traction control means specially adapted for particular types of vehicles for rail vehicles
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B60VEHICLES IN GENERAL
    • B60LPROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L3/00Electric devices on electrically-propelled vehicles for safety purposes; Monitoring operating variables, e.g. speed, deceleration or energy consumption
    • B60L3/10Indicating wheel slip ; Correction of wheel slip
    • B60L3/102Indicating wheel slip ; Correction of wheel slip of individual wheels
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B60VEHICLES IN GENERAL
    • B60LPROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L2200/00Type of vehicles
    • B60L2200/26Rail vehicles
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00Road transport of goods or passengers
    • Y02T10/60Other road transportation technologies with climate change mitigation effect
    • Y02T10/64Electric machine technologies in electromobility

Abstract

LOCOMOTIVE WHEELSLIP CONTROL SYSTEM

ABSTRACT OF THE DISCLOSURE

A locomotive propulsion system includes wheelslip control with the following interrelated features: comparing the rotational speeds of the locomotive's separately driven wheels and detecting the highest and lowest; providing a reference speed value indicative of the rotational speed of a wheel that is not slipping; determining a desired maximum difference between the reference speed and a selected one of the detected speeds; deriving a speed error value by summing the reference speed, selected speed, and maximum difference; obtaining a wheelslip correction value which is a function of the speed error value; reducing the magnitude of voltage applied to the locomotive's traction motors in accordance with the correction value.

Description

-l- 20-LC-1532 LOCOMOTIVE WHEELSLIP CONTROL SYSTEM

~ 9 ~41 ~o_J~ L~ion This invention relates generally to fontro7. syste~s for electrically propelling traction vehicles such as diesel-electric or stra7ght electric locomotives~ and it relates more particularly to improved means for controlling such a vehicl~ in a manner that avoids 05 or minimizes undesirabl~ slippage of the vehicle wheels when "motoring"
(propuls~on ~ode of operat~on) or "braking" (dynamic retarda~ion mode of operation).
Modern locomotives and other large self-propelled tract~on vehicles commonly have, per vehicle, at least four axle-wheel sets l0 (each set comprising a pair of wheels affixed to opposite ends of a rotatably mounted axle), with each axle-wheel set being connec~ed via suitable gearing to the shaft of a separa~e electric motor commonly referred to as a traction motor. In the motoring mode of operation, the traction motors are supplied with electric current fro0 a lS controllable source of electric power (e.g., an engine-driven traction alternator) and apply torque to the vehicle wheels which exert tangential foree or tract7ve effort on the surface on which the vehicle is traveling (e.g., the parallel steel rails of a railroad track), thereby propelling the vehicle in a desired direction along the right 20 of way. Alternatively, in an electrical braking mode of opera~ion, the motors serve as axle-drtven electrica1 generators; torque is applied to their shafts by the~r respectively ass~eiated axle-wheel sets wh~ch then exert braking e~fort on th~ surface, thereby retarding or slowing the vehicle's progress. In either case, good adhesion between each 25 wheel and the surface is required for e~f~cient operation of the veh~cle.
It ls well known that ~aximum tract~ve or braking e ff ort is obtained ~f each powered whe@l of ~he veh~cle is rota~ing at such an angular velocity that ~ts actual peripheral speed is slightly htgher 30 (motoring) or slightly lower (braking~ than the true veh~cle speed (t.e., th2 linear speed at which the veh1cle is ~ravel~ng, usually referr2d to as nground sp~ed" or "track speedn). The difference between wheel speed and track speed is re~erred to as "slip sp~ed."
Thers ts a relat1vely low limit value of sl1p speed at which peak ~
~`

;. , . :
...
. . , 1 ~26520
-2- 20LC-1532 tractive or braking effort is realized. Th~s value, commonly known as the maximum "creep speed," is a variable that depends on track speed and rail conditions. So long as the maximum creep speed is not exceeded, slip speed ~s normal and the vehicle will operate in a stable 05 ~icroslip or creep mode.
As a practical matter ther~ are times when, du~ to a deterioration of rail conditions or to ~n untoward increase in torque, wheel-to-rail adhesion tends to be reduced or lost an~ some or all o~ th~ Yeh~cle wheels will sl~p excess~vely, i.e., the actual slip speed will be gr~ater than the maximum creep speed. Such a wheelslip condit1On, wh k h is characterized in ~he motoring mode by one or more spinning axle-wheel sets and in the ~raking mode by one or more sliding or skidding axle-wheel se~s, can cause accelerated wheel wear~ rail damage, high mechanieal stresses in the drive components of the propulsion system, and an undesirable decrease of tractive ~or braking) effort.
Many different systems are disclosed ln the relevant prior art for automatically detecting and recoYering from undesirable wheelslip conditions, or for preventing such con~i~ions in ~he firs~ place. See for example U.S. p~tent Nos. 3,437,89~ and 3,728,5g~. V~rtually all wheelsl~p control or correction systems temporarily reduce or ~derate~
the tractlon power in order to cure or to avoid a wheelsl;p condit1On.
During the period of time ~hat such power reduction is in eff~ct, the productivity of the vehicle is undesirably reduced, and for this reason both the amount of derat~on and its duration should be minimized.
In normal motoring operation, ~he propuls~on system of a dtesel-electr k loco00t~ve is so controlled as to estab~sh a balanced steady-stat~ cond~t~on wherein the engine-driven alternator produces9 for each d~screte pos~tton of a throttle handle9 a substa~tiallY
constant, optimu~ amount of e-lectrical power for the tract~on ~otors.
In practlce suitable ~eans ~re provided ~or overriding normal operatton of the propulsion controls and reduc~ng engine load tn response to certain abnormal condit~ons, such as loss of wheel adhesion or a load exceed~n~ the power capabili~y of ~he engine at whatever engine speed the throttle is co~mand~ng. Th~s response, generally referred to as deratton, reduces trac~on power, thereby help~ng the locomot~ve recover from such temporary cond~t~ons and/or prevent1ng serious damage .
., - ,.; ~ :

:
, . ~
. . ~.
-3 - 2 OLC--153 2 to the engine.
In add~tion, the propulsion control system conYentionally includes means for limiting or reducing alternator output voltage as necessary to keep the magnitude of this voltage and the ~agnitude o~ load current 05 from respectively exceeding predetermined sa~e maxi~um levels or limits. 6urrent limit is effective when the locomotive is aecelerating from rest. At low lo~omot1ve speeds, the traction motor armatures are rotating slowly, so their baok EMF is low. A low alternator voltage can now produce maximum motor eurrent which in turn produces ~he hlgh 10 tractive effort required for accelerat1On. On the other hand, the alternator voltage magnitude must be held constant at its maximum level whenever locomotive speed is high. At high speeds ~he traction motor armatures are rotating rapidly and have a high back EMF, and the alternator voltage must then be high to produce the required load 15 current.
Some of the prior art wheelslip control systems attempt to curtail undesirable wheel slippage by monitoring slip speed and oYerriding the normal propulsion controls so as to reduoe traction power if and when the slip speed e~ceeds a permissible limit which depends on the 20 magnitude of motor current (and hence total motor torque). See U.S.
patent No. 4,463,289 and British patent No. 1,246,053. Such creep regulators are advantageous because they minimize the duration and severity of wheelsl~p conditions, thereby enhancing the useful life of the vehicle wheels and the productivity of the vehicle.
Most prior art wheelslip controls use the difference between the rotational sp~ed of the fastest axle-wheel set and the rotational speed of th~ slowest axle-whe~l set to indicate slip speed. Th~s differenoe ls a true measure of slip speed so long as a~ least one axle-wheel set is not slipping and at least one set ~s. But if all sets have the same 30 speed, the d~f~erence speed will be zero and Sherefore cannot indlc~te whether the veh~cle is operat~ng properly, with all of its powered wheels rotating at maximum cre~p speed~ or ~s e~speriencing a "synchronous" wheelslip condit~on, with the sl jp speeds of all axle-wh~el sets s;multaneously exceed~ng the max1~um creep speed.
35 Yarious pr~or art solut1Ons to the probl~m o~ de~ecting synchronous slips are disclosed in U.S. patent Nos. 3,210,630; 4,0759538; and
4,588,932, respect1vely.

.

, Many prior art wheelslip control systems are operative in response to rapid changes of wheel speed. A wheelslip condition is indicated if the rate at which the rotational speed o~ an axle-~heel set is changing increases above a predetermined threshold rate that somewhat exceeds os the maximum acceleration ~in the motoring mode) or deceleration (in the braking mode) obta1ned during normal speed changes of the vehicle.
See, for example, U.S. patent Nos. 3,541,406; 3,482,887; and 4,035,~98.
Th~ creep or sl;p of each powered wheel of the vehicle can be accurately detected by comparing the speed of the axle-wheel set with the actual ground speed of the veh kle. But thts requires equipping the vehicle with suitable means for detecting ground speed such as a radar unit (which is relatively expensive and requires continuous recalibration to avoid errors due to vibration or wheel wear) or an 15 extra, unpowered wheel or roller (which is cumbersome and unreliable).
Another known technique for aYoiding wheelslip during motoring is to control or regulate the voltage applied to the traction motors sn as to hold maximu~ creep speed. In U.5. pa~en~ No. 3,982,164, the maximum voltage limit of an open-loop propulsion control sys~em is set a progra~med amount in excess of the synthesized voltage of a non-sl ipping motor, which progr2~med amount represents the desired maximum percent slip and is automatically adjusted to maximize the cnmbined motor currents and hence total mstor torque. AlternatiYely, as is shown and described in Canadian patent No. 950,559, a 25 closed-loop Yoltage regulator will tend to arrest incipient wheelslipst and the control loop can also include means ~r reducing the voltage set point ~and hence motor speed) in inullediatc response to the onset of a wheelslip condition as indlkated by excess~ve slip speed (see page 29 of the referenced Canadian patent).

S~ar~_o~ the Invent~on A general obJectiYe of the present ~nvention is to proY~de an improved wheelsl~p control system for an electrically propell~d traction veh~cle.
Another objective is to prov~de wheelslip control characterized by its very fast response to the incept~on of an undes~rable wheelsl~p 35 cond~tion over a wide range of track speed and by ~ts effect~ve .

, . , , . . . .

.

-~- 20~C-1532 suppression of such a condition with a relatively small reduction of traotive ~or braking) effort.
A ~ore specific objective of ~he invention is to provide, for a locomotive propulsion system, wheelslip control means that is effective 05 to initiate deration in response to an incipient synchronous slip condition, thereby obviating the provision of an instrument to measure track speed.
A further objectiYe is the prcvis~on, in a wheelslip control system that is compatible with the voltage ~speed) regulation approach 10 of the prior art, of improved means for ensuring a quiek, smooth and stable recovery from a wheelslip condition.
The improved wheelslip control means is useful in a traction vehicle propulsion system comprising a controllable source of electric power for energizing a plurality vf traction motors which respectively 15 drive the wheels of the vehicle, means for varying the magnitude of output voltage (or current) of the source in accordance with a variable control signal, and a con~roller for varying the control signal as neeessary to minimize any difference between a feedback value representative of the actual magnitude of the output voltage (or 20 c~rrent) and a reference signal value which normally depends on the power setting of the vehicle throttle (or brak~ handle). A plurality of suitable speed sensors are respectively associated with the separately driven wheels of the vehicle~
In earrying out the invention in one form, the speed sensors are 25 connected to first means for comparing the rotational speeds of the separately driven wheels and for de~ect~ng the highest and lowest speeds, respect~vely. In association with such first means, second ~eans provides a re~erence sp@ed value indicak~ve of the rotat~onal speed of a vehicle wheel tha~ ~s not sl~pping, and thlrd ~eans 30 determines a normally des~red~ maximum allawable difference between the reference speed and the highest (or lowest) speed detected by the first ~eans. The wheelslip control means also ineludes sum~ing means for deriving a speed error va~ue representat~ve of the algebraic sum of the h~ghest (or lowest) speed, the reference speed, and the aforesaid 35 maximum allowable difference. So long as the h~ghest (or lowest) speed has a des~red value, the speed error value will be zero, but 1f it d~fers from the desired value ~n a predetermined dlrection (e.g., if .. .. . . ~ ... ..

;; ,, . .. . ... .
r . .

the highest speed were greater than desired), a posit;ve speed error value is derived and an integrator is operative to obtain a wheelslip correetion value that increases at an average rate that is a function of the speed error value. Means is provided for reduGing the aforesaid 05 reference signal value by an amount corresponding to the correotion value, thereby correspondingly derat~ng the propulsion system.
For optimum perfnrmance of the propulsion system during any wheelslip condition, the wheelsllp control means should cure the wheelslip with the least possible deration ~a~ount and duration). In 10 one aspect of the invention, this is accomplished by supplying the integrator with a voltage error value related to the speed error value by a variable gain that is a predetermined function of the reference speed value, whereby the wheelslip correction value is representative of the time integral of the vo7tage error value. The variable gain 15 increases from a predetermined minimu~ liMit as the reFerence speed value increases from zero, and preferably it varies in accordance with approximately the second power of the reference speed so that the wheelslip correction value will cause just enough deration at any vehicle speed to promptly cure a wheelsl;p condi~ion. At relatively 20 low track speeds, the correction value will have negligible "overshoot," and the averaye tractive or braking effort will be kept as high as possi~le over ~he entire speed range. (Note ~hat if the gain were constant it would have to be sufficiently high to obtain the required decrement of tractive effort per unit of vol~age ~speed]
25 reduct~on wh~le the vehicle is traveling at high speeds, but such a h~gh gain would be larger than needed at lower speeds when the rat1O of tract1On motor current [and thus torque~ to voltage ~s higher.) Also for optimum system performance, the aforesaid th1rd means of the wheelslip cnntrol means ~s operative to vary th~ normally desired 30 ~aximum d~fference speed as a functlon of the reference speed indicated by the second means. In another aspect of the inv~ntion9 the th~rd means includes means for preventlng the ~aximum d~fference speed from decreasing below a certain minimum te.g., in a range from O.S to 1.0 mph~ when the reference speed ~s relatlvely low, and as the reference 35 speed ~ncreases ~rom a predetermined first value (e.g., 10 mph) to a second, h~gher value (e.g,~ 17 mph3, the third means ls effective to vary ~he max~mu~ d~ference speed as a decreasing percentage of .", . : ~ :

~ `
.

i 1 32~520 reference speed. When the vehicle is traveling at speeds above said second value the maximum difference speed tracks the re~erence speed.
Consequently, the normally desired maximum difference speed changes ; smoothly, as the vehicle accelerates, from a relatively high percentage 05 of reference speed (e.g., approximately 12% at said first value of vehicle speed) to a desirably low percentage (e.g., in the range of 4%
to 8% at speeds above said second value). The minimum limit that is imposed on the maximum difference speed when the veh~cle is traveling at a very low speed will prevent unnecessary derations ~hat might otherwise be caused by spurious speed error values as the vehicle speed approaches zero.
~ hen the speed error value is no longer posit;ve (indicating that the wheelslip condition has been cured), tractive effort should be restored as quickly as possible. In yet another aspect of the invention, the integrator is operat~ve when the wheelslip correction value is greater than zero and the highest speed is not greater than desired (or the lowest speed is not less than desired) to decrease the correction value to zero at an average rate tha~ is independent of the speed error but dependent on the reference speed. As the correctYon value decreases, the value of the aforesaid reference signal rises, and $herefore tractive (or braking) effort is restored. Preferably the rate of decrease varies with the reference speed so long as the reference speed exceeds a predetermined relatively low value (e.g., 7 mph). The relationship between ~he recovery rate and the reference speed is so selected that power will be restored at a desired constant rate ~e.y., 50 HP per second per powered axle~ so long as the Yehicle is traveling at speeds faster ~han the above-mentioned predetermined low value. At any speed slower than ~h~s value, the whe~lslip correct1On value is deereased at a predeterm~ned minimum rate (e.g., oorresponding to 0.75 volts per second at the power source); there~ore power will be restored at a rate that varies ~ith speed, whlle motor current (and hence tractive or braking effort) ~ncreases at a desired, sub~tantlally constant rate. Note, however, that at relatlvely low vehicle sp~eds the current li~i~ mode o~ opera~ion of the propulsion system will normally determine the amount of tract~on power in any event.

. . .

" -.. , ;
, . ' ~ :: ' : :
:
., In a different aspect of the invention, performance of the wheelslip conerol means is improved by proYiding fourth means for respec~ively deriving the rates o~ change of rotational speeds of the separately driven wheels of the vehicle, fifth means for comparing 05 these rates and for providing a rate value representative of the highest rate of change, bistable Ura~e trip" means operative from a reset state to a set state in response to the rate value increasing to at least a predetermined p~ckup level, and addi~ional means effective when the rate trip means is set for supplying a variable value that 10 dapends on either the rate valu~ or the aforesaid feedback value or the product of the two. Seeond summing means is provided for combining the variable value and the integrator output so that the aforesaid wheelslip correction value will actually vary with their sum.
Consequently the correct~on value will abruptly increase whenever the 15 highest rate atta~ns the pickup level of the rate trip means, and the size of the increment increases as the rate value and/or the feedback value ~e.g., ~otnr voltage~ increase. The aforesaid pickup level is relatively low so that the rate trip means will effectively respond to incipient wheelslip conditions when the wheel-surface adhesion of the 20 vehicle is low, as indicated by relatively low total motor current.
However, at higher levels of adhesion this pickup level could be so sensitive as to cause spurious derations. Accordingly, ln an optional feature of the inventi~n the pickup level tracks the adhesion level as the latter varies between predeter~ined values, whereby the rate at 25 which the rate trip means becomes effective is smaller at relati~ely low adhesion levels ~e.g., 1 mph per second at an adhesion level of 10%) than at higher adhesion levels (e~g., 3 mph per seeond at 15%
adhesion).
In-another aspect of the inventlon, responsiveness ts relatively 30 large speed error values is improved by providing means operativ2 if the speed error exceeds a predeterm~ned lev~l for obtaining a "proport~onal" Yalue r~lated to the voltage error value by a coefficient that increases from zero as ~he speed error increases above the predetermined level, and the proportlonal value is also supplied to the second summing means where it is added to the wheelsl~p correctlon value.

, . . . .
.
. :
.~

In a significant aspect of th~ invention, ~he means ~or reducing the aforesatd reference signal value comprises memory means effective while the wheelslip correction value is greater than zero for saving a value approxlmately the same as the referenc:e value at the mo~ent of 05 time when either the speed error value becomes positlve or the rate trip means becomes effective. So long as the memory means is effectiYe, the reference signal value is reduced to a level equal to the saved value minus the wheelslip correction value regardless of the power sett1ng of the throttle. In this manner the degre2 of reference 10 signal correctlon ~i.e., the amoun~ of traction p~wer reductlon or deration) during a wheelslip condition is referenced to the control system para~eters as existing at the time when the wheelslip condition was first detected, and once this condition is cured and the wheelslip correctlon value resets to zero, the reference signal value is returned 15 to its former value. This promotes a smooth and stable transition from wheelslip control to normal control of the propulsion system. Assuming that the vehicle did not gain speed during the wheelslip condition in motoring (or did not lose speed during ~he wheelslip condition ;n braking), the propulsion control system can resume normal operation 20 without rese~ting any of its various reference values.
The wheelslip control ~eans includes means for detecting the number of separately driven wheels of the vehicle that are slipping, as indic2ted by excessively high derivatives of ~he rates of change of rotational speeds derived by the aforesaid fourth means. In another 25 aspect of the invention, the aforesaid second means is so arranged that the reference speed value normally varies w~th ~he rotat~onal speed of a wheel that is not slipping ~the slowest wheel during motor~ng or the fastest wheel during braking) but will not change appreciably whenever the nu~ber of slipping wheels equals or exceeds a predeterm~ned l~mit 30 (e.g., 8 wheels ~n a 12-wheel, 6-axle locomot~ve~. As a result, the reference speed will no~ deviate greatly from a value corresponding to the Yehicle speed if most or all o~ the wheels are slipping, as would be true if a synchronous sl~p were to occur. Aftsr the nu~ber of sl~pping wheels attains th~s limit, any deviat1On caused by ~ehlcle speed actually increas~ng in motoring ~or decreasing in brak1ng) ls soon corrected, because the aforesaid spe~d error value ~ill then increase ~n a positive d~rect1On due to the increasing highest speed . . .. . . .. t ., , .: . ~ . .- : .

: .
, - l0- 2 OLC-153 2 (or decreasing lowest speed in braking~ and the integrator consequently.
increases the wheelslip correction value which will derate traction power so as to reduce the number of slipping wheels below the limit, whereupon the second means will again be able t~ raise (or lower) the 05 reference speed, if and as necessary, to a value corresponding to the new speed of the vehicle.
In yet another aspec~ of the invention, a synchronous slip condition is avoided by providlng means for changing the desired value of the highest ~or lowest) wheel speed by a variable amount that inereases from zero with the length of time that the number of slipping wheels is above the aforesaid predetermined limi~, such change tending to increase the speed error value in a positive direction. Preferably the variable amount increases at a rate that varies with the maximum difference speed determined by ~he aforesaid third means and has an lS upper l;mit that is at least approximately twice the maximum difference speed, whereby the speed error value will increase more rapidly at relatively high track speeds than zt lower speeds. As a result, as soon as the number of slipping wheels exceeds the aforesaid limit, the speed error value is boosted ~o a value tha~ soon exceeds the aforesaid predetermined level at which the means for obtaining the proportional value is operative, and the wheelslip eorrection value is quickly and sufficiently increased to derate trac~ion power before all of the re~aining non-slipping wheels begin to slip. In other w~rds, the desired-value changing means of the wheelslip oontrol means is effective to keep at least one of the vehicle's axle-wheel sets not slipp~ng.
Preferably the wheelslip control means also includes means associated with the aforesaid fourth means for detecting the lowest one of the rates of change of wheel speed, and second bis~able means operat~ve from a reset to a set state in response to ~he lowest rate increasing to at least the aforesa~d p~ekup ~eYel. The prev1Ously mentioned second means is so arranged that the reference speed value will not change appreciably if the seeond bistable means is set, as would be true if all of the powered wheels were changing speeds at excessiYely high rates, and the desir2d-value changing means ~s so arranged as to increase the speed error value by a variable amount that increases from zero with the length of time tha~ the s~cond bistable .
- .

.. " , , . . :

t 326520 means is in its set state. With this arrangement the desired deration can be alternatively initiated either by the second bistab1e means changing to its set state in response to a relatively rapid onset of a synchronous slip condition or by the number of slipping wheels becoming 05 excessively large in response to such a condition developing more gradual ly.
The invention will be better understood and its many ob~ectives and advantages will be more fully apprecia~ed fronl ~h~ following description taken in conjunction with the acoompanying drawings.

Brief DescriDtion of the Drawin~s I0 FI6. 1 is a schematic diagram of the principal components of a propulsion system for a 6-axle locomotive, including a ~hermal prime mover (such as a diesel engine), a traction alternator, a plurality of traction motors and their respectively associated speed sensors, and a controll~r;
lS FIG. 2 is an expanded block diagram of the controller (shown as a single block in FIG. I~ which produces an output s;gnal ~or controlling the field excitation of the traction altsrnator;
FI6. 3 is a diagram of an "equivalent circuit" that is used to illustrate the manner in which the alternator field excita~ion control signal is produced by the controller shown in FIG~ 2;
FIG. 4 is an expanded block diagram of the signal processor ~shown as a single block in FI~. 3) to which the speed sensors of FIG. I are coupled;
FIG. 5 is an expanded block d~agram of the principal components of 25 the whPelslip control means shown as a single block ln FIG. 3;
FI6. 6 is a functional block diagram of the means for computing the values of the speed and voltage errors ir ~he FIG. 5 ~heelslip control;
FI6. 7 ls a flow chart that expla~ns the presently pre~erred 30 manner of generat~ng the reference speed value and the synchronous sl~p tlming function that are ut11ized in o~h~r components o~ the wheelslip control shown ~n FI6 5;
FIG. 8 ~s a block diagram of the integra~ing and proportional functions of the FI~. S wheelslip control;

. .

.: ' -l2- 20LC-1532 FIG. 9 is an expanded block diagram of the rate function shown as a single block in FIG. 5;
FIGSo 13-12 are flow charts that explain the preferred manner of implementing the integrating, proportional and rate functions shown 05 schematically in FIGS. 8 and 9, with FIG. 11 illustrating th2 deration subroutine and FIG. 12 illustrating the recovlery subroutine;
: FIG. 13 is an expanded block d;agram ol a simplified version of the power reference value reduc~ng means shown as a single block in FIG. 5; and FI6. 14 is a flow chart explaining the preferred manner of imple~enting the referPnce value reduction function shown schematically in FIG. 13.

Detailed Description~ of the Invention - The propulsion system shown in FIG. 1 includes a variable-speed prime mover 11 m~chanically coupled to the rotor of a dynamoelectric machine 12 comprising a 3-phase alternating curren~ (a-c) synchronous generator, also referred to as the main traction alternator. The 3-phase voltages generated by the main alternator 12 are applied to a-c input terminals of at least one 3-phase, double-way uncontrolled power recti~ier bridge 13. The direct curr~nt (d-c) ou~put of the bridge 13 is el~ctrically coupled, via a d-c bus 14, in energizing relationship to a plurality of adjustable speed d-c traction motors M1 through M6.
Prime mover 11, alternator 12, ~nd rectifier bridge 13 are suitably mounted on the platform of a self-propelled tract~on vehicle 10 which typic~lly is a 6-axle dies~l-el~ctric locomot~ve. The locomotivc platfor~ is in turn supported on two trucks 20 and 30, the f1rst truck 20 having three axle-wheel sets 21, 22 and 23 and the oth~r truek 30 also having three axle-wheel sets 31, 32 and 33.
Eath axle-wheel set of the locomotive comprises a pair of flanged wheels aff~xed to opposite ends of an axle. All s~x pa~rs of wheels travel on a surfaGe provided by a pair of parallel, spaced-apart rails, one o~ ~hich ~s illustrated by the line 15 in FIG. 1. Each of the traction motors M1-M6 is hung on a separa~e axle, and its armature is mechanically coupled, via convent~onal gear~ng ~not shown1, in dr~ving relationship to the associated axle-wheel set. During ~he ~otoring or 35 propulsion mod~ of operation~ the fi~ld windings of the tract~on motors ~: :

, ~ 326520 Ml-M6 are connected in series with the respective armature wind;ngs thereof, and the six motors are electrically connected in parallel with one another. Suitabl~e current transducers 2~ are used to proYide a family of six current feedback signals II through I6 that are 05 respectively representative of the magnitudes of mo~or armature current, and suitable speed sensors 28 are used to provide a ~amily of six speed feedback signals Wl through W6 that are respPctively representative of the rotational speeds (revolutions per ~inute, or ~rpm"~ of the motor shafts and hence of the s~parately dr~ven axle-wheel sets.
The maîn alternator 12 and the reotifier bridge 13 serve as a controllable source of electric power for the respectiYe traction motors. The magnitude of output voltage (or current) of this source is determined and varied by the amount of excitation current supplied to field windings 12F on the rotor of the main alternator. The alternator field excitation current is supplied by a 3-phase controlled rectifier bridge 16 the input terminals of which receive al~ernating voltages from a prime mover-driven auxiliary alternator 18 tha~ can actually comprise an auxiliary set of 3-phase windings on the same frame as the main alternator 12. Conventional control or regulating means 17 is provided for varying the magnitude of direct current that the controll~d rectifier bridge 16 supplies to the alternator field (and hence the output of the alternator 12) in accordance with a variable control signal VC on an input line 19. The control signal YC is provided by a controller 26 which, as will be more fully explained when FI6. 3 is de~cribed hereinafter, is operative to vary VC as necessary to minimize any difference be~ween a re~erence signal (the value of wh kh normally depends on the value of a variable command signalJ and a feedback siynal representat~ve o~ the actual value of the quant~ty being regulated. The system includes su~tabl0 means for derivlng a voltage feedback signal V representat~ve o~ the average magnitude of the rectified output voltage of the ma~n alternator, Which magnitude is a known functton of the field current magnitude and also varies with the speed of the prime mover 11.
The prime mover 11 ~s a thermal or ~nternal-combustion engine or equ~valent. On a diesel-electric locomo~ive, the mo~ve power is typically provided by a high-horsepower, turbocharged. 16-cyllnd2r .

:, . .

diesel engine. Such an engine has a fuel system ~ot shown) that includes a pair sf ~uel pump racks for controlling how much fuel oil flows into each cylinder each time an associated fuel injector is actuated by a corresponding fuel cam on the engine crankshafts. The 05 position of each fuel rack, and hence the quiantity of fuel supplied to the engine, is controlled by an output piston of an engine speed governor 35. The governor regulates engine speed by automatically displacing the racks, within predetermined I;mits, in a direction and by an amount that minimizes any difference between actual and deslred 10 speeds of the engine crankshaft. The desired speed is set by a variable speed call signal received from the controller 26~
In a nor~al motoring or propulsion mode of operation, the value of the engine speed call signal is determined by the position of a handle of a manually operated throttle 3~ to which the controller 26 is coupled. A locomotive throttle conventional ly has eight power posit10ns or notches (N), plus idle and shutdown. N1 corresponds to a minimum desired engine speed ~power), while N8 corresponds to maxi~um speed and ~ull power. In a consist of two or more locomotives, only the lead unit is usually attended, and the controller onboard each trail unit will receive, over a trainline 37, an encoded signal that indicates the throttle position selected by the operator in the lead unit .
For each power level of the engine there is a corresponding desired load. The controller 26 is suitably arranged to translate the 25 throttle nctch information into a reference signal value substantially equal to the Yalue that the voltage feedback signal Y will have when the ~ract~on power matches the called-~or power, and so long as the alternator output voltage and load current are both w~thin predetenm~ned lim~ts the control signal VC on the input l~ne 1g of the 30 field controller 17 1s varied as neoessary to obtain this desired lo3d.
For thts purpose, and for the purpose of derat~on (i.e., unloading the eng~ne) and/or limiting englne speed in the event of certain abnormal conditlons, it is necessary to supply the controller 26 ~ith infor~ation about Yarious operat1ng conditions and parameters of the 35 propuls~on system, includ~ng the engine.
As 1s tllustrated in FIG. 1, the eontroller 26 receives voltage ~eedback signal V, current feedback signals Il-I6, axle/wheel speed .

, . , , feedback signals ~l-W6, another current feedback signal representative of the magnitude of the rec~ified output current of the main alternator 12, an engine speed signal RPM indicating the rotational speed of the engine crankshaft1 a load control signal issued 05 by the governor 35 if the engine cannot develop the power demanded and still maintain the called-for speed, and relevan~ data from ather selected inputs 38. (The load control signal is effective, when lssued, to reduce the value of the reference signal in th~ oontroller 26 so as to weaken the alternator field until a new balanc~ point is reach~d.) In an electrical braking or retarding mode of operation, inertia of the moving vehicle is converted into elec~rical energy by utilizing the traction motors as generators. To configure the propulsion system ~or brakiny, the throttle handle is moved to its idle posit~on, an IS interlocking handle of a brake control device 39 is moved from an off position, thrcugh a set-up positlon, tu various on positions, and the armature windings of the traction motors M1-M6 are disconnected from the motor field windings and reconnected to an appropriate load circuit wh~ch, in the case of dynamic braking, typ kally comprises an array of fan-blown resistor grids (not shown3 where electrical energy generated by the motors is dissipated in the form of heat. The motor field windings are now separately excited by the rectified output current of the main alternator 12, and the controller 26 is operative to vary the alternator field excitation so that at relatively hi~h track speeds (e.g-, h~gher than approximately 29 mph) the average magnitude of ~otor current supplied to the resistor grids (and hence braking effort) is regulated to a reference value that depends on the setting of the brake handle, whereas at lower track speeds~ wh~never ~axi~um ra~ed current ~s reached in the motor field windlngs, the alternator output current is held constant at thls limit and braking current will decrease linearly with speed.
In the presently preferred embodiment of the invention, the controller 26 comprises a microco~puter. P~rsons skilled in the ar~
will understand that a microcomputer is actually a coordinat~d system of cowmercially available components and assoc~ated electrical circuits and ~lements that can be programmed to perform a variety of des~red ~unct~ons. In a typical microcomputer, which is illustrated in FIG. 2, . .
.
.
.
~. . .
. : , !
: ,' , a central processing unit (CPU) executes an operating program stored in an eras~ble and electrically reprogrammable read only memory (EPROM) which also stores tables and data utilized in the program. Contained within the CPU are conventional counters 9 registers~ accumulators, o5 ~lipflops (flays), ete., along with a precision osoillator which provides a high-frequency clock signal. The microcomputer also includes a random access me~ory (RAM) ln~o which data ~ay be temporarily stored and from which data may b~ read at various address locations determined by ~he program stored ~n the EPROM. These components are interconneeted by appropriate address, data, and control buscs. In one practical embodimcnt of the invent~on, an Intel 8086 microproeessor is used.
The other blooks shown in FI~. 2 represent conventional peripheral and interface components that in~erconnect the microcompu~er and the external circuits of FIG. l. More particularly, the block labeled "I/O" is an input/output circuit for supplying the microcomputer with data representative of the selected thro~tle (or brake) positlon and with digital signals representative of the readings of various voltage, current and other sensors associated with the locomotive propulsion syste~. The latter signals are derived from an analog-to-digital converter 41 connected via a conventional multiplexer 42 to a plurality of signal conditioners ~o which the sensor outputs are respectively applied. The signal conditioners serve the conventional dual purposes of buffering and biasing ~he analog sensor output signals. As is indicated in FI6, 29 the input/output cirouit also interconnects the microcomputer with the engine speed governor, the eng1ne speed sensor, th~ axle/wheel speed sensors, and a d~gital-to-analog signal converter 43 whose output VC is connected via the line 1g to the alternator field controll~r.
The controller 26 is progra~med to produce, in the motoring mod*
of operation, a control signal having a value that varies as necessary to zero any error between the value of the alternator voltage feedback signal V and the value of a refer~nce signal that normally d~pends on the throttle position selected by the loeomot~ve operator and the traction power output of th~ ~ain alternator. The manner in which th~s is aceompl1shed is functlonally illustrated in F~6. 3. In addition~ in order to lmplemcnt an electrical braking mode o~ operation, th2 .

~ .

.

controller can be programmed to vary the valu of the control signal as necessary to zero any error between thg value of ~he motor armature current feedback signal (i.e., the average malgnitude of the six current feedback signals I1-I6) and the value of a re~erenee signal that 05 normally depends on ~he dynamic brak2 pos~t~on selected by the locomottve operator.
The throttle positlon da~a ~ha~ the con~roller 26 receives in motoring is hereinafter referred to as a var~able command signal; it is fed, as is shown in FIfi. 3, to a block 45 t~at represents su1table means for p~rforming decoding, deration, and rate limit Functions.
Preferably these functions ar@ carried out ~n the manner disclosed in United states Pa~nt Nuuiber 4,634,887 issued January 6, 1987, to Balch et al, and assigned to General Electric CompanyO The bloc~ 45 has first and second outpùt channQls, labeled "PWR" and "V~I," respectively. A number representing a desired value of traction power output of the main alternator per powered axle of the locomotive is provided on the ~irst channel, and a number representing a desired value o~ the voltag~ (and current) limit is provided on the second channel. Under normal, steady state operat1n~ conditions, suoh desired values are determined by the Yalue of the com~and signal. Bu~ when more power is called for, the rate li~it function in the bl~ck 45 will cause the desired data to change at a controlled rate which is a function o~ the actual speed and horsepower of the engine. ~
As can be seen in F~G. 3, ~he desired power value on th~ fjrst output channel of the block 45 is supplied as one input to first summing means 46 via a block 47 whose funct~on is to impose a preset maximu~ limlt on th~ des~re~ power va~ue. Th~ datum representln9 the desired voltage ~and curr~nt~ lim~t on the second output channel ~s suppl~ed as one tnput to second summing means 48 via a block 49 whose funct~on is to establ1sh an absolute maximum llm~t for the altern~tor output current, and the same datu~ is also deployed as a Yoltage limit reference value wh~ch ~s separately suppl~ed, via 2 block 50 and a llne 51, as one lnput to third summing ~eans 52. The funct~on of the block 50 1s to establish an absolut~ maximum limit for the alternator output voltage. The th~rd summing means 52 has anoth~r inpu~ ~h~ch is th~
value of the voltage feedback signal V. The latter valu~ represents .

"~
..

1 32h520 - l8- 2 OLC--153 2 the average magnitude of the rectified output voltage of the main alternator, and the third summing means derives, on a line 53, a voltage error value representing the amount that this magnitude is under the desired voltage limit.
os The second summing means 48 has another input which is the value of a current feedback signal IMAX derived from a current processor 55 to which the six current feedback signals Il-I6 are supplied. The processor 55 is so constructed and arranged that IMAX is determined by whichever one o~ the ind~vidual current feedback signals is largest, and thQrefore the value of IMAX corresponds to the highest ~agnitude of motor curren~. The second sum~ing means 48 derives a difference value representing the amount that the highest current magnitude is under the desired current limit. This difference value is proeessed in accordance with a programmed system compensation routine 56 ("System Comp I") to derive a current limit reference value that is swpplied on a line 57 as one ;nput to fourth summing means 58. Th~ fourth summing means has another input which is the value of the voltage feedback signal Y, and it derives, on a line 59, a curren~ error value equal to th~ reference value on the line 57 minus the value of V.
As is shown in FIG. 3, the first summing means 46 has another input which is the value of a power feedb~ck signal HP proportional to the product of the matn alternator output voltage and the traction motur current. To obtain the power feedback value, the value of the voltage feedback signal V is multiplied, in a block 61, by the value of a current feedback sign~l IAV which is derived in the current processor 55 and whtch is representative of the total mc~or current divid~d by s~x. The result1n~ signal KVA is represen~at~v~ of the k1lowatts of tract~on power, and ~t {s divlded, in a block 62, by a scaling factor to provide th~ power feedback signal HP corresponding to tractian : 30 horsepower per powered axle. The flrst summi~g means 46 derives adifference value represent~ng the amoun~ (if any) that the actual traction power is less than desired. Th~s difference value is processed in accordance with another program~ed system oompensation rout~ne 63 to derive a power reference s~gnal that is supplied, ~ia a 3S line 64, a wheelslip control function 65, and another l~ne 66, as one input to fifth summing means 67~ Associa~ed with the wheelslip control ~unction 65 ls a signal processor 68 coupled ~o the six ~ndiv1dual . ... ... ...

. -; , -axlz/wheel speed sensors 28. In a manner that will be fully explained hereinafter, the signal processor 68 and the wheelslip control function 65 are operative in response to incipient or actual wheelslip conditions to override the throttle-dependent reference signal on the 05 line 64 and reduce the power reference value on the line 66, thereby derating traction power in order to restore wheel-rail adhesion.
The fi~th summing means 67 has another input which is the value of the voltage feedback signal V, and it der1ves, on a line 69, a power error value equal to the reference value on the line 66 minus the value of V. The voltage, current, and pow~r error values on the lines 53, 59 and 69 are respectively processed by programmed error compensation routines 71, 72, and 73 ("Error Comp") to derive voltage, current, and power control values. The error compensation routines 71-73, as well as the system compensation rout~nes 56 and 63, introduce proportional-plus-integral transfer functions the respective gains of which are functions of the throttle position and other para~eters of the locomotiYe and its controls, as determined in a gains function 7~.
Thus each of the compensated control values varies as a function of the time integral of its associated error value. All three of the control values are supplied as inputs to a least value gate 74, and the value of the control signal at the output of the gate 74 is the smallest of these input values. The resulting control signal determines the magnitude of the analog eontrol signal VC that the controllDr 26 applies, via the line 19, to the alternator field controller 17 (FIG.
~5 1).
The field controller 17 will respond to the control signal value by varying the field strength of the tract~on alternator 12 as necessary to ~in~mize whichever one of the error values on the lines 53, 59, and 69 is producing the smallest input value to the least v2lue gate 74. So long as both V and IMQX are with~n the l~m~ts set by the oommon V ~ I output of the block 45 and are not above their respective maxi~um lim~ts as establ~shed at blocks 50 and 49, the control s~gnal value is determined by the power con~rol value wh1ch will now be s~aller than either the voltage or current control value.
Consequently9 the alternator output voltage is ~aintained at whateYer level results in the value of V equall~ng the power reference value wh~ch normally i~ the same as the value that the system compensation .

routine 63 es~abl i shes on the 1 i ne 64 when there is no error between the sensed magnitude of traction power and the ~esired magnitude thereo~. But if V ~or IMAX) tends ~o exceed its limit, the voltage (or current3 control value is driven lower than the power control value and 05 the control signal value accordingly decreases, whereby the alternator voltage is adjusted to whatever level resul$s in essentially zero error between V ~or IMAX) and the voltage (or current) limit reference value.
As was explained earlier, FI&. 3 shows the manner o~ controlling the ~otoring operation of the loco~ot~ve propulsion system.
Alternatively, the same controller can control the electrical braking or retarding operation of the propulsion system. In the la~er case, the variable command signal would be the brake handle position data, the value on the first output channel o~ the block 45 would represent the desired value of brake current (i.e., the magnitude of current in the resistor grids per powered axle), the feedback signal for the first I summing means 46 would be the motor armature current feedback stgnal i IAV, and the feedback signal for the third, fourth and fifth summing means would be the alternator output curren~ feedback signal I (i.e., the magnitude of current in the traction motor field windings~.
The speed signal processor S8 and the wheelslip con~rol function 65, comprising the wheelslip control ~eans of the present invention, are illustrated in more detail in FIGS. 4 and 5, respectively. As is shown functionally in FIG. 4, the speed feedback signals Wl-W6 are fed to a suitable calibration function 76 which scales each of these rotat~onal spesd signals to the corresponding peripheral speed value (miles per hour, or "mph~) of the associated axle-wheel set and which from time to time adJusts these values as necess2ry to ensure that all six of them are substanttally equal to one another when the locomot~ve is coasting (1.~., when the traction motors Ml^M~ are deenergized and track speed exceeds 8 mph) despite unequal wheel d~ameters. The calibrated speed feedback signals are ~n ~urn fed to suitable ~eans 77 for comparing their values and for detecting the highest and lowest values, respectively. The latter ~eans 77 has two output values: nne 1s labeled nWMAX~ and varies with the highest speed detected, and th~
cther is labeled ~WMIN~ and varies with the lowest spe~d det~cted.
The six calibrated speed feedback signals ~rom the cal1brat~on function 76 are also fed to suitable f~rst differentlat~ng means 80 for .

.

~' , 1 32652~
-2l- 20LC-1532 deriving six rate of change of speed values. ~ate values ~hat have a positive sign are representative of the acceleration (mph per second) of the respectively associated axle-wheel sets, whereas relatively negative rate values are representative of the deceleration (mph per 05 second) of the respectively associated axle-wheel sets. All six rate values are in turn supplied to suitable means 81 far comparing them and for detecting the highest and lowest values, respect1vely. The latter means 81 has two output values: one is labeled "dW/dtMAX" and varies with the highest rate of change o~ speed, and the other is labeled ~dW/dtMIN~ and Yarles with the lowest rate of change of speed.
The six individual rage values derived by the first differentiating means 80 are also supplied to suitable second differentiating means 82 for deriving six absolute values, herein called "jerk" values. These jerk values are derivatives of the respective rate values and therefore are second derivatiYes of the callbrated speed feedback values. In other words, each jerk value is representatlve of the rate of change of acceleration ~or deceleration) - expressed in units of mph per second per second - of the associated axle-wheel set. As is shown in FIG. 4, the jerk values are supplied to the inputs of six bistable level detecting means 83 where they are respectively compared with a preseleoted threshold level provided by a block 84.
The output state of each of the level detectors 83 is deter~ined by whether or not the associated jerk value exceeds the preselected 25 threshold level which pr~ferably is s~ightly above the maximum rate of change of acceleration ~or decelerat~on) expected during normal speed changes of the locomot~ve with good wheel-rail adhesion~ In one practical application o~ the ~nvent~on, the threshold level is 0.75 mph/sec~; if any jerk value exceeds ~his level ths associated 30 axle-wheel set is assumed to be sp~nn~ng or sk~dding. The level detector outputs are mon~ored by log~c means 85 for detecting the number of separately dr~ven axle-wheel sets tha~ are sllpping, as tndicalted by the output state of each 1 evel detector wh~n the associated jerk value is excessively h~gh. Preferably, the logic means 35 85 is so arranged that its outpu~ count (lab~l~d "#") equals the number of a~le-wheel sets ~ha~; are not sl~pping, as ~nd~ca~ed by ~erk values not exceeding the aforesatd ~hreshold level. In practtce ~he processor 68 also includes appropria~e low-pass filtering means (not shown) for suitably conditioning the input and/or output values of various components thereof.
As is shown functionally in FI~. 5, the highest and lowest speed 05 values, the highest and lowes~ rate valuec~ and ~he non-slipping axle count from the above-described speed signal proc2ssor 68 are fed to the wheelslip control function 65, along with the voltage feedback value ~V), the average current value tIAV) d~rived from the current processor 55, th~ power reference value derived from th~ system compensation block 63, the current li~it reference value der1ved frQm the system compensat~on block 56, and the voltage limit reference value on the line 51. The wheelslip control function 65 comprises five coordinated subsyste~s plus a plurality of collateral logic steps for initiating the alternative deration and recovery modes of wheelslip operation. In describing the presently preferred embodi~ent of the subsystems and logic steps of the wheelslip control function, it will be assumed that either constant or increasing traction power is being called for by the command signal. As will be apparent to persons skilled in the art, by suitably interchanging certain of the illustra~ed inputs and/or their polarities the invention would also be useful, if desired, to control wheelslip during electrical braking.
In the first subsyste~ 100 of the wheelslip control function 65, a speed error value (DeltaW) and a voltage error value ~De~taV~ are computed from th~ known highest speed value (WMAX1 and a reference speed Yalue generated by the second subsystem 200. The reference peed value is indicative of the rotational speed of a 10comotive axle-wheel set that is not slipping, and it is therefore a true ~easure of the track speed of ths locomotive (except under a synchronous wheelslip condit~onj. As will soon b~ explained ~hen FI6. 6 is described9 the speed error value is representative of the algebraic sum of the h~gh0st axle-wheel speed, the reference speed, and a predeter~ined normally desired maximum difference spe2d. If the highest speed has a desired value, the speed error value is zero. Whenever ~he highest speed has an actual value greater than desired, the speed error value is positlve. Upon detecting a posit~ve speed error ~+D~ltaW), the f~rst subsystem 100 causes an OR log~c step 87 to have a h~gh output state that in turn act~vates a deration or power reducing mode ot' operation ' ' .
.

of the wheelslip control function. As is illustra~ed functionally in FIG. 5, the output of a second OR logic step ~ will be in a high or "TRIP" state concurrently with the high state of the step 87.
In the deration mode of operation, the first subsystem 100 is 05 operative to derive a voltage error v~lue related to the speed error value by a variable gain that is a predeter~ined func~ion of the reference speed value. The voltage error value is then integrated in the third subsystem 300 of the wheelslip control funct~vn 65. This produces a variable value "INT" that is represen~at~ve of the time integral of the voltage error value, increas~ng in a positive sense at an avera~e rate that varies with the voltage error value.
Oonsequently, the INT value increases at a rate that is a desired function of the speed error value. It is fed via a first output line 301 of the subsystem 300 to summing means 90. So long as INT is positiYe (+INT), the subsystem 300 activates one input of an AND logic step 91.
If and when the speed error value is positive and exceeds a predetermined level, the third subsystem 300 is operative to produce a second variable value "PROP~ proportional to the voltage error value.
2~ The PROP value is fed via another output line 302 to the summing means 90 which derives a wheelslip correction value that varies with the sum of the INT value, the PROP value, and a third variable value "RATE."
The latter value is fed to ~he summing means via a line 401 from the fourth subsystem 400 whenever the highes~ rate (dW/dtMAX) equals or exceeds ~ predetermined pickup level, and it will raise ~he wheelslip correct~on value by an amount that var~es with the product of the hlghest rate value and the value of ~he vol~age feedback signal Y. If desired, the pickup level can be a function of the whee7-rail adhesion of the locomotlve, as indicated by the average value ~IAY) of motor 3~ current; for example, ~e piekup level could vary from one mph per second at an adhesion level of 10X to thre~ mph per second at 15%
adhesion~ Once the highest rate attains the pickup level, a RATE TRIP
is indicated and th~ subsystem 400 causes ~he logic s~ep 87 to activate the deration mode of operat~nn o~ the wheelsl~p control funct10n. At the same time the log~c step 8B is in lts TRIP state.
The wheelsl~p correct~on value derived by the su~ming means 90 ~s an input to the fi~th subsyste~ 500 of the wheelsl~p control function.

."
, ., ~ , .

1 32h520 ~henever the logic step 88 is in i~s TRIP s~ate, the latter subsystem is operative to override the throttle-dependent power reference signal received on ~ts input line 64. In operation, as will be explained below when FIGS. 13 and 1~ are described, t~is subsystem will reduce os the value of the power reference signal on its output line 66 by an amount equal to the correction value~ So long as the reduced reference value on line 66 is lower than the power reference value on the line 6~1, the wheelslip control funct~on 6~ ;s in fact in control (WS
INCONTRL) and the subsystem 530 act~vates the other inpu~ of the AND
logic step 91. Only if both of its inputs are concurrently activated (~.e., WS INCONTRL and +INT coexist), the step 91 will have a high output state.
As is illustrated ~unctionally in FIG. 5, the logic step 88 will be in its TRIP state concurrently with the high output state of the logic step 91. An addit~onal logic step 92 is associated with the two log k steps B7 and sa; it has a high output state that activates a recovery or power restoring mode o~ operation of the wheelslip control function 65 whenever the output of step 87 is not high ~i.e., the highest speed is not greater th n desired and the deration mode has terminated) while the step 88 is in its TRIP state (i.e., ~he INT value is posit~ve and hence the wheelslip correction value is greater than zero, and the wheelslip control function is in control). In the recovery mode of operation, ~he third subsystem 300 is effective to decrease the INT value, and hence the wheelslip correction value, at an average rate that is ~ndependent of the speed error value and that preferably varies as a predetermined ~unction of ~he reierence speed.
As soon as the INT value ~s returned ko zero, bo~h o~ the logic StBpS
91 and 88 are deactivated and thus change from h~gh to low st3tes. Nnw the wheelsl~p correct~on vaTue is zero and the recovery mode of operation is term1nated. At the same time~ via broken llne 94, the th1rd subsyste~ 300 re-in~t~al~zes the values of the system compensation represented by the block 63 in FIG. 3. After the recovery mode terminates, the wheelsl~p correction value derived by the summing means 90 will remain equal to zero, the power reference value on the l~ne 66 will be the same as the value on th~ line 64, and the propulsion control syst0m can operate ~n its normal mo~oring mode unless and until another derat~on mode ~s in1tlated by the concurrent .. . . ..

:, ~
;. , ,~
.

.
~, low-to-high state changes of the logic steps 87 and 88 ln respanse to either a positive speed error being detected or a RATE TRIP being indicated.
With reference now to FIG. 6, the first subsystem 100 of the 05 wheelslip control function 65 will be more partieularly described.
Persons skllled in the art will understand that in praetice th~
functions shown schematically in FI~. 6 are best implemented by suitably programming the microcomputer that comprises the preferred embodiment of the controller 26. The speed error value (DeltaW) is computed by al~ebraically combining, at a sum~ing point or st~p 102, four different input values. The first input is the previously described h~ghest speed value (WMAX) taken from on~ output of the detecting ~unction 7~ ;n the speed signal processor 68 (FIG. 4). The second input is obtained from a least-value gate 104 which selects the smaller of two values: the reference speed value provided by the referenee speed generator 200, and the lowest speed value (WMIN) taken from the other output of the detecting function 77. Although the reference speed value is usually equal to or smaller than WMIN, it may be larger under some circumstances, (e.g., while the locomotive is decelerating in its motoring mode). The second input is subtract~d from the first input.
The third input to the sum~ing point 102 is the aforesaid maximum difference speed value. It is set at a variable peroentage of the reference speed value. As is illustrated symbolically in FIG. 6, the reference speed value is reduced to the desired percentage by a mult~plying step 106 the output of which provides one input to a greatest-value gate 108. The ga~e 108 has a second Input esupled t~ a ~ntmu~ level block 110 that supplies a relatively low, constant value, and it seleets the larger of its two inputs as the value that is suppl~ed, via a line 112, to the summ~ny point 102. The latter value is representative of the maximum ereep speed, i.e., the normally des~red maximum difference betwe~n the reference speed and WMAX. The purpose of the greatest-Yalu~ gate 108 is to prevent the maximum dtfferenc~ from decreasing below a predeter~ined minimum speed (e.g., in a range from approximately 0.5 to 1.0 ~ph) when the reference speed ~s relat~vely low (e.g., less than 8 ~ph3.

' :

The outpu~ of the multiplying step 10~ is thP product of the reference speed value and a variable ~raction which is also a ~unct1On of th2 referenee speed value, as determined by a function yenerator 114. Preferably, the generator 114 is operative ~o vary this fraction 05 ;nversely with the re~erence speed as the reference speed varies between a predetermined firs~ value (e.g., approximately 10 mph) and a predetermined second value (e.g., approximately 17 mph). For exa~ple, the fraction is variable between a maximum limi~ of approximately 1/8 gso that the maxi~um d;fference speed would ectual 12% of the referenee speed i~ reference speed were lower than its first value but higher than the speed at which the output of the multiplying step 106 is the same as the low value supplied by the minimum level bloek 110) a~d a minimum limit in a range of approximately 1/25 to 1/12 (so that the ~aximum difference speed is not less than 4X of the reference speed so long as the reference speed is higher than its second value). As a result, the maximum difference speed will be a desirably higher percentage of reference speed when track speed is low than when traek is relatively high, and the percentage decreases smoothly from maximum to ~inimum limits as the locomotive accelerates. As is shown in FIGc 6, the maximum difference speed value on ~he line 112 ~i.e., the third input to the summing point 102~ is subtracted from WMAX (i.eO, the first input).
ThQ ~ourth input to ~he summing point 102 is a Yariable valus equal to the product of two input values of a multiplying function 116.
The first input (TIME) is a variab~e faetor ~aken from a synchronous slip timer in the reference speed genera~or 200 of the wheelslip control function 65. As will soon be explained when FI6. 7 1s described, this fae~or is normally zero, but it will increase at a linear rate from 0 to a maximum of 1.0 ~f the synchronous sl~p timer were activate~ 1n response to an incipient synchronous slip condit~on being indlcated by either of ~wo even~s: ~1) the output count of the logic means 85 (FI6. 4~ ls less than a eertain a~ount, thus ind~cat~ng that th~ number of sl~pping axle-wheel sets exceeds a desired l~mit; or ~2) the lowest rate value (dW/dtMIN) equals or exceeds a predetermined pickup level, thus indicating that all six axle-wheel se~s are changing speeds at excess~vely high rates.

.

,: ,, : . :

The second input to the mul~iplying function 116 is proportional ta the normally desired maximum difference speed; i~ is obtained by multiplying, in a step 118, the maximum difference value on the line 112 and a predetermined number (at least approximately 2, and 05 preferably 3). The resulting product of th~s multiplying step is a chosen multiple of the maximum difference speed value, and it provldes one input to a ~reatest-value ga~e 120. The latter gat~ has anoth~r input coupled to a minimu~ lev~l block 122 supplying a constant value corresponding to a predetermined minimum speed (such as approximately one mph), and ~t selects the larger one of its inputs as the value that is supplied as the second input of the mult~plying function 116. The greatest-value gate 120 will prevent the second input fro~ decreasing below a minimum level corresponding to a predetermined minimum speed when the maxi~um difference speed is lower.
It will now be apparent that the output of the multiplying function 116 is 0 so long as its TIME input is 0, but when the synchronous slip timer is activated the valu2 of this output starts increasing toward an upper limit equal to the chosen multiple of the maximum diff~rence speed value. Consequently, the fourth inpu~ to the summing point 102 is a variable value that increases with the length of t~me that the synchronous slip timer is activated. The rate of increas~ will vary with the maximum difference speed value cn the line 112 ~but cannot fall below a minimum rat~ determined by ~he minimum level 122), whereby the fourth input increases more rapidly at relatively high trac~ speeds than at lower speeds.
The fourth input to the summing point 102 is added to the first ~nput ~WMAX), thereby tending to increase or boost, in a posit~ve direction, the speed error value (DeltaW~ obtained at ~he output of this sumnling point. In effect, the four~h input reduces the a~oresaid desired value of the hlghest axle-whe~l speed by a var~able amount corresponding to the value of the fourth ~nput~ In other words, th@
desired speed equals the sum of the reference speed plus the normally desired max~mum d~ference speed ~inus said variable amount. In operatlon, if some bu~ not all of the axle-wheel sets w~re actually sl1pping and the synchronous slip timer were not ac~vated, the value o~ ~MAX would be greater than the su~ of the reference speed value plus the maximum d~fference speed value. Alternatively, if enough . .

..
. .

1 3265~0 -~8- 20LC-1532 .
axle-wheel sets were slipping ~or if all sets were accelerating fast enough) to aetivate the synchronous slip timer? the sum of the reference speed value plus the maximum difference speed value minus the variable value of the fourth input to ~he summing point 102 would 05 become less than the value of WMAX. In either event, the actual highest speed would be greater than desired, and the speed error value would have a positive sign and a magnitude determined by the size of the deviation from desired speed.
The speed error value derived from the summing point 1b2 in the subsystem 100 is tested by a suitable bistable polarity detecting means or step 124, and whenever it is positive (+DeltaW) the dete ting step 124 is operative in conjunction with the logic step 87 (FIG. 5) to activate the deration mode of operation of the wheelslip control function and the TRIP state of the logic step 88 as described hereinbefore. During the deration mode of operation, the speed error value is mult~plied by a variable gain that is a predetermined function of the reference speed value to derive the voltage error value ~DeltaVI. In FIG. 6 the multiplying step is symbolically illustrated by a block 126 labeled UX~ having first and second inputs, with the speed error value being supplied directly to the first input and the reference speed value being coupled via a function generator 128 and a normally open switch 130 to the second 1nput. Whenever the deration mode is in effect, the switch 130 îs clnsed and the multiplying function will be operative; otherwise the voltage error value is zero.
The function generator 128 sets the value of the gain (i.e.t the ratio of the voltage error value to the speed error value~ that is supplied to the second input of the multiplying step 126 when operat~ve. It is programmed so that ~he gain ~"GAIN") has a predetermined minimum limit ~K~ (e.g., 1.0) and ~ncreasss in a non-linear manner to a predetermined 3~ max~mu~ limit (e.g., 22.5~ as ~he reference speed increases to a relatively hi~h value te.g., 70 mph). Thus, GAIN ~s dependent on reference speed, and preferably it Yaries in accordarlce with approximately the second power o~ the re~erence speed value so that over a wlde range of track spe~ds the voltage error value will be just large eno~gh to cause trac~ve ef~ort to decrease a predetermined desired amount (corresponding, for example, to a reduction in total traction motor current of 200 amps) in response to a speed error of .

, -29-~ 20LC-1532 1 ~ph. As a result, the corrective ac~lon of the wheelslip control function will be sufficient to cure any wheelslip condition that occurs at high track speeds (whcn the ratio of motor current ~and thus torque~
to voltage is relatively low) without undesirable overshoot if a OS wheelslip oecurs at low traek speeds. In a pr~ctlcal embodiment of th~
invention the functions explained above in this paragraph are actually implemented by the software shown in FI6. 10 and described below.
FI6. 7 ~llustrates the presently preferred way to implement the reference speed generating and synehronous slip tim~ng functions represented by the block 200 in FIG. 5. The controller 26 is suitably programmed to execute the FI60 7 routine 50 t~mes a second. The routine starts at an inquiry point 202 which de~er~ines whether or not the locomot~ve throttle 36 is being moved from its idle position to its first power position (N1). The answer to this inquiry will be true whenever the operator is co~manding the locomotive to accelerate from rest. (In the flow charts, Y stands for yes or an affirmative answer to an inquiry, and N stands for no or a negative answer.) If the answer to the first inquiry 202 were aff~rmative, the FIG.
7 routine would i~mediately execute a referen~e speed initializing step 2~ 204 before it proceeds to the next inquiry point 206. In the step ~04, a temporary register o~ the microcomputer (herein referred to as the ~reference speed" register) is loaded with the value of WMIN ~in binary form) which represents the rotational speed of the slowest axle-wheel set as detectPd by the speed signal processor 68. The next inquiry 206 tests whether or not WMIN is less than the value stored ~n the reference speed register. If affirmative (indicating that the locomot~ve is decelerattng), the routine proceeds fro~ inquiry 206 to a step 208 ~h~ch is used? if necessary, to ensure that two b1stable flags ~m~ntmu~ raten and ~maximum rate~ respectlYely~ are ~n first or reset states (here~n referred to as ~urned off or ~falsen). After execut~ng the step 298, the FI6. 7 rout~ne automatlcally proceeds ~o a step 210 where a delta value is calcula~ed by subtracting WMIN from the reference speed valu~ and ~ultiplying the difference by a desired gain ~REF-DN~). The st2p 210 ~s followed by an inquiry step 212 ~o determ~ne whether or not the calculated delta value exceeds a preset slew lim~t: ~f true, ~he del~a value ~s reduce~ to the value of the slew limit at a step 21~. This ~s automat~cally followed by an inquiry .
,; "", I
~, .

,: ' , ` 1 3~65~0 _~o_ 20LC-1532 216 to determine if the delta value is less than the reference speed value. If so, a step 218 is e~ecuted to decrement or reduce the value stored in the reference speed register by an amount equal to the delta value; otherwisej the stored reference speed value is reduced to zero os by a step 220. In either event, the reference speed generating routine ends here.
Alternatively, if the locomo~ivs were either aceelerating or not changing speed, or if there is an actual or incipient wheelslip c~ndition, the answer to the inquiry 206 would be no, and the FIG. 7 routine would then proceed from this inquiry point to a decision step 222. The latter step de~ermines whether or not the nu~ber (#) of non-slipping axle-wheel sets is below a predetermined limit "LIM"
(e.g., 2 for a locomotive having six powered axle-wheel sets): if #
were below LIM, an affirmative answer would be obtained at the step 22Z
and the routine would proceed directly from 222 to a timer-actlvating step 224; otherwise the next step to be executed would be another inquiry 226. In other words, an affirmative answer at the decision step 222 indicates that the number of slipping axle-wheel sets is above a desired maximum li~it (e.g., 4). In this event, as was explalned hereinbefore, a synchronous sl~p eondition is imminent, and now a counter will be incremented by a predetermined number, up to a certa~n maximum count, at the step 224. The eounter-incrementing step is the synchronous slip timer, and it is the last step of the reference speed generating routine when the decision step 222 is aff1rmative. Ths count stored in the eounter determines the aforesaid TIME lnput to the multiply~ng funct1On I16 in ghe subsystem 100 (see FI&. 6~, and the maximum count corresponds to 2 TIME factor of 1Ø So long ~s # ~s below LIM, the count will be period~cally increased every time the FIG.
7 rout~ne is executed, each increment being e~ual ~o the predetermin~d number unt~l the maximum count is reaehed. By wa~ of example, the predetermined number can be 1/250 of the maximum count, whereby an interYal of five seconds is required for the TIME fac~or to ~ompl~te tts max~mum excursion from 0 to 1.0~ Thus the count stored 1n the counter is a ~easure of khe t~me tha~ the number o~ slipping axle-wheel s2ts 1s above the desired l~m~t.
The counter-increment~ng step 224 is also activated every time the FI6. 7 rout~ne is executed when the number of slipping axle-wheel s~ts ..
.. . .
,, . ~
" ... . .

~ . . .

-" ~ 32~520 is not above the desired limit (i.e., # is not below LIM~ if the lowest rate value dW/dtMIN then equals or exceeds the prede~ermined pickup level (nAH) that indicates a relatively rapid onse~ of a synchronous slip condition. As is shown, when there is a negative answer at the 05 decision step 272 the next step 226 of the rou~ine is to test the state of the minimum rate flag. So long as this flag is no~ in its second or set state ~herein referred to as turned on or "truen), the routine proceeds via a step 228 to an inquiry 2~0 which determines if the lowest rate of change of speed detected by thP sp~ed s~gnal processor 10 68 is at least as hi~h as A. While not shown in FI6. 7, means is provided for varying A as a function of the magnitude of traction motor current. Only if the answer to the inquiry 230 is yes, the counter-incrementing step 224 is activated via a step 232 that changes the setting of the bistable minimu~ rate flag from false to ~rue. Once the steps 232 and 224 are executed, the FIG. 7 routine ends, but the next time through this routine the answer to the inquiry 226 will be yes instead of no. Now an inquiry is made at a step 234 to determine whether or not the lowest rate value is less than the dropout level "Bn which is a predetermined faction of A. If not, the latter s~ep is inlmediately followed by the counter-increment~ng step 224.
Alternatively~ if and when dW/dtMIN d~creases below B, the routine will automatically proceed from the step 234 to the step 228 wh@re the bistable minimum rate flag is rese~ to its false state, and from the step 228 via the inquiry step 230 to another decision step 236 which deter~ines whether or not the number of non-slipping axle-wheel sets equals th~ afor~said predetermined amount. If the answer at the latter step were affirmative, the FIG. 7 routine would end here; otherwise the numb~r of sllpping axle-wheel sets is below the desired limit, and the rout1ne will n~w proceed from the decis~on step 23~ to a s~ep ~38 that resets the aforesaid counter ~1.e., the synchronous slip timer3 to 1ts normal state of z~ro.
After the counter is reset at the step 23~, the next step ?40 of the FIG. 7 rout1ne is to t~st ~he state of the maximum rate ~lag. As long as th~s flag is not in ~s second or s~t sta~e (herein referred to as turned on or ~true"), the routine proceeds via a step 242 to an inquiry 244 which determines i~ the hlghest rate value d~/dtMAX is at least as high a~ a predeterm~ned pickup level wh~ch preferably is the .
' ' ' .~

~ :
, -32- 2t)LC-1532 same as the aforesaid pickup level A. If the answer to the inquiry 244 ls yes, the FIG. 7 routine ends after a final step 246 that ehanges the setting of the bistable maximum rate flag from false to true. The next time through this routine after the step 246 has been executed, the 05 answer to the inquiry 240 will be yes lnstead uf no. Now an inquiry 248 is ~ade to determine whether or nat the highest rate value is less than a predetermined dropout level (preferably the same as B3. If nct, the latter step concludes th* FIG. 7 routine. Alterna~vely, if and when dW/dtMAX decreases below B the rout~ne will automatically proceed from the step 248 to the step 242 where the bistable maxi~um rate ~lag is reset to its false state. From the latter step the routine prooeeds via the ~nquiry 244 (the answer to which is now negative, since B is lower than A) to the next step 250. Note that if the answer to inquiry 206 were to change from no to yes at any time after step 246 sets the maximu~ rate flag in its true state but before this flag is reset to its false state by the step 242 ~or after the mini~u~ rate flag is set in its true state by step 232 but before it 1s reset by s~ep 228), the rate flag would be reset by the step 208.
In the step 250 a delta value is calculated by subtracting the reference speed value from WMIN and multiplying the difference by a desired gain ("REF-UP"). Th~s step is followed by an inquiry step 252 where ~he calculated delta value is compared with a preset slew limit:
if the.c!alculated value exceeds the limit, it is reduced to the value of the slew limit at a step 254. This is automatically follnwed by a step 256 that increments or increases ~he value s~ored in the reference speed register by an amount equal to the delta value, and ghe reference speed generat1ng routine ends here.
The operation of the reference speed genera~or ~llustrated in FIG.
7 will ;now be briefly summarized. During normal decelerat~on of the locomot~ve, the answer to the inqu~ry 206 is aff~rmatiY2 and the steps 208 ?20 will be effect~ve to e~sure that ~he reference speed value tracks the decreasing lowest speed value WMIN, with the rate of r~ference speed ohange correspon~ing ~o ~he rate at wh k h WMIN is varying up to a maximum rate determined by the preset slew l~it.
During normal acceleratlon of a loeomotive, the answer to the ~nquiry 206 is negative and the steps 222, 226, 228, 230, 23~, 238, 240, 242, 244, and 250-256 will be effeotive to ensure tha~ the reference speed ~ ~ , .

-` 1 326520 33_ 20~C-1532 value tracks the inoreasing WMIN, with the rate of reference speed change again corresponding to the rate at which WMIN is varying up to the same maximum rate. In either case, the reference speed value will vary with the rotational speed o~ whichever one of the axle-wheel sets 05 has the lowest speed. That one set normally is not slipping, and its speed is therefore an accurate ~easure of the aotual traok speed.
However, the illustrated routine will not appreciably change the reference speed value whenever any of the fo710wing abnormal events is detected~ # is equal to or less than LIH, as indlcated by an affirmative answer at either one of the decision steps 236 and 222 (i.e., when the number of slipping axl~-wheel sets equals or exceeds the desir~d maximum limit); (2) dW/d~MIN increases to at least A and then remains above B, as indicated by the minimum rate flag being set in its true state (i.e., when the lowest rate value is abnormally high); or ~3) dW/dtMAX increases to at least A and then remains above B, as indieated by the maximum rate flag being set in its true state (i.e., when the highest rate value is excessively high~. The first or second of these three events can occur when most or all of the axle-wheel sets begin to slip, as would be true if a synchronous slip condition were to develop while the locomotive is traveling at a speed in e~cess of approximately 10 to 15 mph, and in either event the reference speed is prevented from deviating greatly from a value corresponding to the actual track speed which tends to remain constant during synohronous slips. The third event can occur when any one of the axle-wheel sets is changing speeds at an excessiYely h~gh rate, as would be true lf at least one set were to slip while the locomot~ve is traveling at a rzlatively low track speed, and ln th1s event the reference speed is preYented from changing fro~ a value corresponding to the actual track speed at the t~nle when the excessive rate is 30 detected so as more closely ~o ma~ch the track speed desired when the systen~ subsequently recovers frolll the wheelsl~p condit~on.
The referenc~ speed generator 200 includes the synchronous slip t~mer (step 224) which is actlvated either by ~;he above-ment~oned dW/dtMIN event or when ~he nu~ber of non-slipping ax1e-wheel sets fa11s 35 below the aforesaid predetQrrined amount or limit. In ei~her case, the step 224 ls executed dur~ng each pass ~hrough th~ FI6. 7 routine, thereby periodlcally ~ncreasing the TIME factor from ~ toward 1.0 at a ~:
.. .
.. ..

t ~6520 _34_ 2 OLC-153 2 linear rate. Consequently, ~he mul~iplying function 116 ~FI6. 6) will now be operatlve to increase its ou~put at a rate that varies with the normally desired maximum difference speed so as correspondingly to boost the speed error value (DeltaW) as previously explained. If th~s 05 synchronous slip timing act~on were initlatecl by # falling below LIM
(c.g., 2), when # subsequently increases to a number equal to LIM
whatever count is then stored ~n the counter (step 224) would be saved.
As a result, the TIME factor is "frozen" until ~ ei~her again falls below such limit ~whereupon an affirmative answ~r i~s again obtained at 10 the decision step 222 and the step 224 will be PffectiYe during each pass through the rout1ne to resu~e incrementing the counter from its saved count toward its maximum count) or rises above such limit (whereupon a negative answer is ohtained at the decision step 236 and the next step 238 then resets the counter to a zero count). In other words, after the number of slipping axle-wheel sets increases abo~e the desired limit (e.g., 4), the synchronous slip timing step 224 is operative periodically to increase the TIME factor that is used as an input to the multiplying function }16, thereby increasing the variable value that the latter function contributes to ~he summing point 102 20 (see FIG. 6~. Subsequently, as the system recovers from a wheelslip condition, the number of slipping axle-wheel sets will decrease. Now the decision step 236 is operative to hold the TIME faetor ~an~ hence the variable value at the output of the multiplying ~unction 116) relatively constant so long as the nu~ber of slipping sets equals the desired limit, whereas the resetting step 238 is operative abruptly to reduce this factor (and hence the var~able value) to zero in response to the number of slipping sets decreasing ~o below such limit. As soon as the synchronous slip t~mer is thus reset and the variable value is thereby reduced to zero, there ~s a corresponding step decrease of the s~eed ~rror value, and th~s event can cause the wheelslip sontrol funct~on immed~ately to sw~tch ~rom derat~on to recovery modes of operat10n. The decis~on step 236 will delay th~s event unt1l the nu~ber of slippjng axle-wheel sets ~alls below the limit~ at wh k h point there is less like~ihood of a borderline synchronous slip oondit10n cau~ing undesirably rap~d cyc7~ng between recovery and derat10n modes of operation.

.

: ;
~, ,. , . ,; . . , ,.
,~ . ,, . ~ , ,, , ` , j ' , - 1 326~20 It will now be apparent ~hat the reference speed generator 200 (FIG. 7) provides both the ref~rence speed value and the TIME factor witho~t using a radar unit or an extra, unpowered whe*l or roller to measure the actual track speed of the veh~ole. In the manner 05 illustrated in FIG. 6 and previously describe~d, the reference speed value is algebraically summed with the highest speed value, the maximum difference speed Yalue at the output of the greatest-value gate 108, and, when the synchronous slip ti~er is active, the variabl~ value at the output of the multiplying function 116 so as ~o derive the speed error value (DeltaW) which ~s mult~plied by a variable gain to derive the aforesaid Yoltage error value (DeltaV). The latter value is integrated in the third subsystem 300 of the wheelslip control function 65. FI6. 8 illustrates the subsystem 300 functionally, it being understood that in practice these functians are best i~plemented by suitably programming the microcomputer comprising the preferred e~bodiment of the controller 26. (Flow charts of the presently preferred software are shown in FIGS. 11 and 12 and will soon be described.) As is illustrated in FIG. 8, the voltage error value is passed through a transfer function or gain 303 ~o one input of a greatest-value gate 304, the other input of which co~prises a pred~termin~d minimll~n value obtained from suitable ~eans 306. The latter means is programmed to change the minimunl value fro~ a first1 relatively low amount to a second, higher a~ount in response to the reference speed value increasing above a predetermined level "K1~ as detected by assooiated level detecting means 3a8. (K1 is attained, for example, at a relatively low rQference speed of 8 mph. ) The greatest-value sate 304 seleots the larger of ~ts two inputs as the value that is supplied, via first and second ser~es-connected switches 30 312 and 314, respecti~/ely, to the input of an integrat~ng function 310.
The first switch 312 1s normally open, but it will be closed whenever the derat~on mode of operation is in effect. The other switch 314 has tWb alternatlve positions or states: normally i~ is in ~h~ state shown ~n FI~. 8 so as to connect the input of the integrator 31û to the 35 switch 312; but whenever the recovery mode of operation is in effect, the sw~tch 31~1 is in its second state, thereby connecting the integrator input to the output of an alternative source 316 that 1s :. : , ...

- , independent of the speed and voltage error values. Suitable means 318 is provided for inverting the polarity of the latter ou~put, whereby the integrator input is negative whenever the switch 314 is in its second state. The source 31~ is suppl~ed with the re~erence speed 05 value, and it is programmed so that its output: varies with ~he latter value so long as the lat~er value exceeds a certain level "K2"
corresponding to ~ predetermined reference speed (e.g., seven mph).
Th~s source is also programmed so that a~ relatively low reference speed values ~less than K2), its output has a preselected minimum valus 10 IlDo t~
The integrator 310 is operative to produce the aforesaid variable value INT on the first output line 301 of ~he subsystem 30Q. During the deration mode of operation, INT will be representative of the time integral of DeltaY. In other words, the integrator output increases at an average rate that varies with the voltage error value. This rate of increase w~ll not be less than a minimum determined by the minimum value obtained from the means 305. During the recovery mode of operation, the alterna~ive source 316 is operative to decrease INT at an average rate that is independen~ of both ~he voltage error value and the speed error value, and that is a function of the reference speed.
More particularly, the rate of INT decrease will vary with the reference speed value so long as the latter value exceeds K2, and it equals a constant minimum rate wheneYer the reference speed value is less than K2. The minimu~ rate (preferably chosen ~o cause the 25 rectified output voltage of the ma~n alternator 12 to deorease at a r2te of appro%imately 0.75 volts per second) ~s detQrmined by the selection of the minimum output value ~ of the alternative sDurce 316.
From the previous description of the logic step 92 ~FII;. 5) ~t will be recal~ed that the recovery mo~e becomes effective whenever the 3~ speed error changes from a posit~ve value to a nega~ve value (or zero~, and it ~s deact~vated as soon as the INT value r~turns to zero and i:s th~refore no longer positive. If the propulsion system were operat~ng in a normal creep-regulating mode~ the wheelslip oontrol function 65 would repeatedly cyele between derat~on and recovery ~odes 3~ as WMAX gradually varies between values ~hat are slightly greater and sllghtly less than the desired speed. D~ring the al~ernate intervals when the speed error value has a positive d~rec~ion or sign9 INT would ", , ~ .
.. . . ;'' ' ' ' ' .
' ', ' ' ' , ' ~, '' ', ~ , ' _37 2 OLC-153 2 -ramp-up ~rom zero and thereby reduce t~e power reference value on the line 66 (FIGS. 3 and S), whereas during the intermittent intervals when the speed error value is not positive, INT would ramp down ~o zero and thereby increase the value on the line 66 which has the effect of 05 restoring traction power.
As is illustrated in FIG..8, the integrator output INT is tested by a suitable bistable polarity detec~ing means or step 320 coupled ~o the line 301, and so long as INT is rela~ively positive (+INT) the detecting step 320 is operative in conjunction with the WS INCONTRL
~0 indicat~on from tha subsystem 500 to cause the output of the AND logic step 91 to be in its high state which keeps the OR logic step 88 in its TRIP state (see FI6. 5). Therea~ter, as soon as ei~her INT has been reduced to zero or the wheelslip control func~ion loses control (i.e., the power reference value on the line 64 becomes lower than the reduced reference valu~ on the line 66), the AND logic step 91 is de-activated and ;ts output state will therefore change from high to low, whereupon a reset function 322 becomes effective to ensure that the output of the integrator 310 is clamped to zero. At ~he same ~ime, a companion reset function 324 is effective to re initialize the sys~e~ compensation values used in the compensation routine 63 (FIG. 3) to which the function 324 is coupled via the line 94.
The third subsystem 300 includes a multiplying function or step 330 for producing the aforesaid variable value PROP on i~s second output line 302. The voltage error value is applied directly to one of the two ~nputs of th7s step 330~ and the speed error value is coupled v~a a function generator 332 to the second input. Th~ funct~on generator 332 is programmed to pr~vide a variable mult~plier that increases from zero to a prede~ermined maxi~um nu~ber (e.g., four) as the sp~ed error value increases above a prede~ermined leYel ~L~ wh~ch corresponds, for example, to a speed error of one ~ph. In other words, whenever DeltaW is hi~her than L, 332 is effect~ve to supply to the second input o~ the multiplying step 330 a multiplier tha~ varies with the difference therebetween. Conse~uently, only when ~he speed error value is posit~ve and exceeds L the multiplying step 330 is operative to produce PROP which ~ill then be related to the voleage error value by a coefficient that varies between zero and the aforesaid max~mum number as the speed error value var~es between L and a predetermined .
., ..:.. ,j secsnd level higher than L. In practice, DeltaW will usually increase to a value exceeding L when the multiplying function 116 is operative to supply the fourth input to the summing point 102 (see FI6. 6) as a result of activation of the synchronous slip timer (step 224 in FIG. 7~
05 in the reference speed generator 200, whereby PROP will rise quickly from zero in response to an incipient or actual synchronous slip condition being indicated at either one of the steps 222 ~nd 230 of the FIG. 7 routine.
As was previously explained, the wheelslip control funct~on 65 includes summing means 90 (FI6. 5) for algebraically combining the v~lues labeled PROP and INT from the subsystem 300 and the value labeled RATE from the subsystem 400 to derive the wheelslip correction value that serves as an input to the fifth subsyste~ 500, whereby the correction value comprises the sum of these three separate values or amounts which are representative, respectively, of the product of DeltaV and a multiplier that varies with the amount by which DeltaW
exceeds L, of the time integral of DeltaV, and of the product of V and dW/dtMAX if the latter exceeds the pickup level A.
With reference now to FIG. 9, the four~h subsyste~ 400 from which zo RATE is obtained will be more particularly described. In practice the ~unctions shown schematically in FI6. 9 are best implemented by suitably programming the microcomputer that comprises the preferred embodiment of the controller 26. The subsystem 400 is arranged to convert the highest rate value (dW/dtMAX), taken from one output of the rate comparing and detecting means 81 in the speed signal processor 68 (FIG. 4), into the variable value RATE which is proportional to the product of ~he absolute ~agnitude o~ dW/dtMAX and the value of Y. For th~s purpose the highest rate value is coupled via a normally open switch 402 to one input of a first multiplying funct~on or step 404 the other lnput of which receives the voltage feedback signal Y, and the resulting product of the first step 404 is ~ultiplied by a predetermined fraction tnX~) in a companion multiplying step 406. tThe two ~ultiplying steps 404 and 406 may be implemented concurrently in praet~ce.) Th0 predetermined fraction 1s obtained fro~ suitable means 408 whlch can be programmed, lf desired, to ~ncrease thls fract~on in response to the reference speed value increasing above a predetermined leYel as detected by an opt~onal level detector 410. The multiplying 1 3~65~0 -39. 20LC-1532 steps 404 and 406 are e~fective when the switch 402 is elosed to produce the aforesaid variable value RATE on the su~put llne 401 of the subsystem 400. The latter value depends on both the highest ra~e value and the value of V, being the predetermined percentage of their Q5 product.
WheneYer a RATE TRIP ~s indicated, the switch 402 ts closed and the multiplying step~ 404 and 406 will be operative as described above;
otherwise ~ATE is zero. The subsystem 400 includes suitable bistable level detecting means responsive to the highest rate value for determining whether or not to indicate a ~ATE TRIP condition. In FTG.
9 this level detecting means is illustrated as a single block 412 which is functionally the same as the steps 240-248 of the FIG. 7 routine.
It has a normal or reset state in which its ou~put is low or "9~ and a set or RATE TRIP indicating state (coinciding to the turned on or "true" state of the maximum rate flag) in which its output is high or ll1." The latter state begins whenever the highest rate value increases to at least the pickup level A, and it terminates whenever the highest rate value subsequently decreases below the dropout level B. As is illustrated in FIG. 9, A is determined by suitable means 414, and B is a predetermined fraction (e.g., one-half) of A as determined by a multiplying step 416. These two levels are fed in tandem to both the level detecting means 412 and the re~erence ~peed generagor 200.
In an optional feature of the invention, the pickup level determining means 414 is a function genera~or suitably programmed to 25 vary A as a predetermined function of the wheel-rail adhesion of th~
loeomotive. The adhesion level is estlmated from the known weight of the locomotive and from th~ magnitude of trac~ivn motor current, as represented by the Yalue of the average current feedback signal IAV
derivsd ~n the current processor 55 (FIG. 3). ThQ value of IAV tends to Yary with the level of adhesion; that is, as adhesion changes, motor current (and hence torque) will change correspond~ngly. The relat~onship bet~een IAV and adhesion for a given weight can be stored in a lookup table of the microcomputer. The function generator 414 is so arranged as to vary A ~ith ~he adhesion level as the adhesion level varies between predetermined first and second percentages (e.g., 10X
and 15%, respectiYely), and ~t ineludes means for determining maximum and ~inlmum l~mits between which A can vary. The pickup level A will , ~ ',~: , . ' '. , .

.

-40 2oLc-ls3z rema1n equal to the minimum limi~ (e.g., one mph per second) when the adhesion level ls lower than the first percentage, i~ will remain squal to the maximum limit te.g., three mph per second) when the adhesion level is higher than the second percentage, and it tracks the adhesion os level inbetween these values. Wi~h A being varied in this manner, the level detecting means 412 will be more sensitive at lo~ adhesion levels (@.9., slippery rails) than at higher adhesion levels (e.g., dry rails).
Whenever the speed of at least one of the axle-wheel sets starts to increase at an excessively high rate while the track speed of the locomotive is relatively low (e.g., less than approxima~Ply 10~15 mph3, the highest rate value attains the pickup level A and the level detectlng means 412 then changes states to indicate a RATE TRIP
condition. The aceeleration at which this state change takes place is less when the adhesion is low and a wheelslip condition is more likely to occur than when adhesion is relatively high. Once the state change takes place, the logic step 87 (FI~. 5) ensures that the deration mode of the wheelslip control function 65 is activated, and the switch 402 ~FIG. 9) is olosed to enable the mult~plying steps 404 and 406 of the subsystem 400 to feed the variable value RATE to the summing means 90 (FI6. 5) as previously explained. The latter value ~which is proportional to the product of the highest rate value and the valu~ of V), will increase the wheelslip correction value at ~he output of the summing means 90, and the power reference value on the line 66 is consequently reduced a corresponding amount so as to prevent a wheelslip condition from occurring or to cure one tha~ has already occurred.
The deration and reoovery modes o~ opera~on will now be reviewed w~th referenc~ to FIGS. 10-12 which arg flow char~s o~ th~ presently pref~rred software for implementing the functions shown schematloally ~n FI6S. 8 and 9. The wheelsl~p ~ode control routine illustrated in FIG. 10 ~s automat kally e%ecuted 50 times per second. It beyins at a first inquiry step 602 (equ~valent to the polarity detect~ng step 124 in FI6. 6) that checks whether or not the speed error value ob~ained at the output of the sum~ing point 102 (FIG. 6) is posit1Ye. If not, a second inquiry step 604 of the rout~ne will check the state of a b~stable "rat~-trip~ flag. Th~s flag is the above-ment~oned max~mum ., rate flag which is changed from a reset or false state ~o a set or true state by the step 246 ln the reference speed generat~ng routine shown in FIG. 7 and which is returned to its reset state by the step 242 or 208, and it is equivalent to ~he level detee~ting means ~12 shown in 05 FIG. 9. Until the rate-trip flag is set in a true or high state, ns RATE TRIP is indicated and the FI6. 10 routine automatically proceeds from the step 604 to a third inquiry step 606 that will soon be described. Alternativ2ly, if the answer at either one of the inquiry steps 602 and 604 were affirmative (equiYalent to the OR logic step 87 10 in FIG- 5 having a high output state), that step would be immediately followed by a step 608 where the setting of a bistable "trip" flag is changed, if this flag were then in its reset or "false" st~te, to ~ set or "true" state (equivalent to the OK logic step 88 in FIG. 5 being in its TRIP state).
The FIG. 10 routine proceeds from the step 608 to a step 610 which reads the values stored in the aforesaid reference speed reg~ster (see description of FIG. 7) and finds GAIN by multiplying the reference speed value by itself and adding the minimum li~it K to the product.
This step includes means for imposing a predetermined maximum limit on GAIN. In the next step 612 the positive speed error value is multiplied by &AIN, and the resulting product is the aforesaid voltage error value which is saved in a temporary register of the microcomputer. It will now be apparent that the steps 610 and 612 are equivalent, respectively, to the function generator 128 and the ~ultiplying step 126 of the subsystem 100, as illustrated in FI6. 6 and described above.
The next step after the step 612 is ~o actiYate a deration subrout~ne 614, and this ls auto~atically followed by exscution o~ a reference value redueing subroutine 616 which cs~pletes the FI6. 10 3~ routlne. During e~ery pass through the routlne, the polarity of the speed error value is recheckPd at the f~rst ~nquiry 602, and ~f ~t is not positive, the rate-trip flag is rechecked at the second 1nquiry 504 to determine if th~s flag is stlll set ~n ~ts true sta~e. Whenever negatiYe answers are obtained at both o~ ~he steps 602 and 604, the FI6. 10 rout~ne proceeds to the thtrd inquiry step 606 instead of to the step 6~8. The state oF the trip flag is tes~ed at the step 606.
If not true, the routtne proceeds directly to the subroutine 616. But 1 ~2652Q
-42~ 2 OLC-153 2 if the tr;p fl3g were now in its true state (eguivalent to the logic step 92 1n FIG. 5 having a high output state), ~he routine would activate a recovery/termination su~routine 61~ before proceeding to the step 616. Details of the presently preferred embodiment of the 05 deration and recovery subroutines 614 and 618 will next be described with reference to FI&S. 11 and 12, respeotively.
As is illustrated in FIG. 11, the deration subroutine is entered at a step 621 where the voltage error value that was saved at the preceding step 612 (FIG. 103 is reduced to a clesired increment or step value (nINCRMNTn) equal to the product of the v~ltage error value and a predetermined integrator gain ~e.g., 0.08). Then the reference speed value is tested at an inquiry point 622 ~o de~ermine whether or not it exceeds the predetermined relatively low level K1. If the answer were yes, the next step 623 would load a temporary register with a predetermined first minimum value 'IHI;'l otherwise a step 624 is operative to load the salne register with a second, lower minimum value IlLOW.l' After the minimum value is set in this mann~r, it is compared with the aforesaid incremental value at a step 625. So long as the latter exceeds the former, the deration subroutine will proceed directly from the step 625 to an integrating step 626; otherwise the step 626 is preceded by an extra step 627 that makes the desired increment equal to the minimum value that was set at the step 623 or 624. It will now be apparent that the s~ep 622 performs the same function as th~ level detecting means 30~, the alternative steps 623 and 624 are equivalent to the minimum value means 306~ and the steps 625 and 627 are equivale~ ~o the gr~at~st-value gate 304 of the subsystem 300, as illustrated in FI6. 8 and described above.
In the integrating step 626 of the deration subrout~ne, the aforesaid increment is added to an ~OLD INT~ value taken from a temporary register of the microcomputer (herein referred to as th~ ~old integrator value" register~, and the~r sum is the previously described variable value INT which increas~s ~n a pos~tive direction at an average rate tha~ vari~s with the voltage error value. The step 626 also reloads th~s sum (in binary form) into the old integrator value register where it ~s saved as OLD INT for th~ next pass thrcugh the FIG. 11 subroutine. In an associated step 628 that follows the integratlng step 626, ~he same sum (INT) ~s lo~ded ~nto another . .

~ . .; , `

~ 326520 _43_ 2 OLC--153 2 temporary register (herein referred to as the "wheelslip correction value" register) where it is saved as the "old correction" value.
After the steps 626 and 628 have been executed, the state of the rate-trip flag is checked at an inquiry point 629. If this flag were os in its set or true state (indicating that the highest rate value dW/dtMAX has attained at leas~ the pickup level A), the subroutine would proceed from the inquiry 629 to a step 630 where dW/dtMAX and the value of the voltage feedback signal Y are retrieved and multiplied and the resulting produot is multiplied by ~he predetermined fraction ~%.~
Thus the step 630 is effeetive to produce th~ previously described variable value RATE that depends on both the highest rate value and the voltage feedback value (see functions 404-408 of the subsyste~ 400 as illustrated in FI6. 9). It is automatically followed by a summing step 631 that adds RATE to the value saved in the wheelslip correction value register, and the re5ulting sum is reloaded into this register where it is saved as the old correction value. It will be apparent that onee the rate-trip flag changes ~rom false to true states, the correction value saved at step 628 is increased when the step 631 is executed during each pass through the deration subroutine, and the size of the increment (RATE) will increase as the product of dW/dtMAX and V
increases.
From the step 631 the deration subroutine proeeeds to another inquiry point 632. Alternatively, if no RATE TRIP were indicated and therefore the answer to the inquiry 629 were negative, the inquiry 632 would immediately follow the inquiry 629. The inquiry point S32 determines whether or not the speed errnr value DeltaW is greater than the predetermined level L. If no~, the next step 633 will set the a~oresaid vari~ble coe~ficient equal t~ zero, and th~s ~s followed by a step 634 w~ere the voltage error value DeltaV is ~ult~plied by the coefficient. The resulting product will be zero so long as the coefficient is zero. Alternatively, if ~he answer to the inquiry 632 were af~irmat~ve, the step 634 would not be executed until after a step 635 wh~ch is programmed to calculate the variable coefficient by subtracting L from DeltaW and multlplying the difference by a predetermined constant F. Thus th~ coef~icient w1ll ~ncrease linearly from zero as DeltaW increases abcve L. The s~ep 635 is followed by an inquiry 636 that tests whether or not the calculated coefficient ~s ,~,., " .

44_ 20LC-1532 less than the aforesaid maximum number ("MAX"~. If the answ~r to this inquiry were affirmative, the subroutine would proceed directly to the step 634; otherwise, a step 637 sets the coefficient equal to MAX
(e.g., four) before proceeding to the step 634. In either event, the 05 step 634 is now effective to produce the previously described variable value PROP that is proportional to both DeltaV and the variable coefficient. (It will be apparent that the steps 632, 633, and 635-637 are equivalent to the function generator 332, and the step 634 is the same as the multiplying step 330 o~ the subsys~em 300, as illustrated in FIG. 8 and described above.) After the step 634 is executed, the FIG. 11 subroutine will return to the wheelslip mode control routine (FIG. 10~ via a final step 638 that adds PROP to the old correction value. The value derived from the step 638 is the previously described wheelslip correction value that corresponds to the sum of INT, RATE and PROP. Thus the two summing steps 631 and 638 of the deration subroutine are functionally equivalent to the summing means 90 shown in FIG. 5. It will be apparent that so long as DeltaW exceeds L, during each pass through this subroutine the final summiny step 638 will effect a s~ep increase of the value saved in the wheelslip oorrectivn value register, and the size of the step or increment (PROP) equals the prsduct of DeltaV and the variable coefficient which in turn depends on the amount by which Delta~ exceeds L up to a predetermined maximum limit.
So long as there is an affirmative answer at either one of the inquiry steps 602 and 604 (FIG. 10), the deration subroutine (FIG. 11) will be executed every pass through the wheelslip mode control routine.
Consequently the aboYe-described integrating step 626 will periodically incr~ment th~ OLD INT value sa~ed in ~he old integrator va~ue register and will save the incre~ented value INT as OLD INT for th2 next pass through. The size of each ~nerement, as de~ermined by the step 621, is usually proportional to the voltage error value, and the number of such increments per second ~s constant (e.g., 503. The s~ep 627 will be operative when the voltage error value is rela~ively low to prevent the increment from decreasing below the minimum size that is set at the step 623 or 624, thereby establishing a minimum rate of IHT increase.
The steps 622-624 will raise or lower this minimum rate in response to the reference speed value increasing above Kl or decreasing below K1, ~ 326520 2oLr-1s32 thereby ensuring that the minimum rate is higher at high track speeds than at reldtively low track speeds.
If the answers at both of the lnqu;ry steps 602 and 604 were ne~ative, and the answer at ~he inquiry step 606 were affirmative (FIG.
05 10), the recovery/terminatio~ subroutine (FIG. 12) would be executed instead of the deration subroutine every pass through the wheelslip mode control rsutine. As is illustrated in FIG. 12, the recovery/termination subroutine is entered at a first inquiry step 641 (equivalent to the polarity detecttng step 320 in FIG. 8) that checks whether or not the OLD INT Yalue saved in the old integrator value register (see the step 626 in FIG. 113 is equal to zers. If no~, a ~econd inquiry step 642 of this subroutine will check the state of a bistable 'iwheelslip in control" flag. So long as there is a WS
INCONTRL indication from the subsystem 500 (see the description of FIGS. 13 and 14 hereinafter), the wheelslip in control flag is set in a true or high state and the answer to the inquiry 642 is no. In this event the FIG. 12 subroutine will automatically proceed from the step 642 to a third inquiry step 643 that will soon be described.
Altern~t~vely, as soon as the answer at either one of the in~ulry steps 641 and 642 changes from negative to affir~ative (equivalent to th~
output state of the AND logic step 91 in FIG. 5 changing ~rom high to low), that inquiry step wsuld be immediately followed by a step 644 ~here a plurality of functions associated with the termination of the recovery mode of operation are performed. For example, the system 25 compensation values are re-initialized (see the description of the reset funotion 324 shown in FI6. 8), and both of the old integrator value and wheelslip correction value reg~sters are loaded with zeros, thereby ensuring that both the OLD INT value and the wheelslip correction value are reset (see the reset function 322 in FI6. 8). In 30 addition, the step 644 is effeetive to re-initialize th~ desired power value on th~ firs~ output channel of the block ~7 (FIG. 3) as necessary to ensure a smooth transition from the recoYery mode of operation Df the wheelslip control means to the ensuing motoring operation of the locomot~ve propulsion sys~em. From the step 644 the FI~. 12 subroutine 35 will return to the wheelslip mode control routine (FI6. 10) via an add~t~onal step 645 where the bistable tr~p flag is changed from its set or true state to a reset or false state ~equival~nt to .
,. . . .
.
.
. .
.. . ... , . ~ . : .

-~6 20LC-1532 d~-activating the logic steps 91 and 88 in FIG. 5). Once the trip flag is thus reset, the inquiry step 606 (FI6. lO) will be effective to bypass the FIG. 12 subroutine during each subsequent pass through the FIG. 10 routine until the above-described step 608 again sets this flag 05 in its true state in response to the occurrence of an actual or ineipient wheelslip condition, as indicated at either step 602 (+DeltaW) or step 604 (dW/dtMAX equals or exceeds A).
During eYery pass through the FI~. 12 subroutine while the trip flag is in its true state and OLD INT is not zero and the wheelslip in control flag is not false, negative answers are obtained at both of the steps 641 and 642 (equivalent to the AND logic step ~1 in FIG. 5 having a high output state). In this event, the subroutine proceeds to the third inquiry step 643 instead of to the step 644. The step 643 tests the reference speed value to determine whether or not it exceeds the a~oresaid level K2. If the answer were yes, the next step 647 would calculate a desired decrement or step value ("DECRMNT") by subtracting K2 from the reference speed value, multiplying the difference by a predetermined constant ~6~l~ and adding the resulting produet to the aforesaid minimum value D. If ~he answer to ~he inquiry step 643 were no, the subroutine would alternatively proceed from this step to a step 648 that makes the desired deeremen~ equal to D. It W7 11 now be apparent that the steps 643, 647 and 648 are equivalent ~o the function generator 316 as illustrated in FIG. 8 and described above.
From either one of the steps 647 and 6~8 the FIG. 12 subroutine proceeds to another inquiry step 649 that determines whether or not the aforesaid decrement is less than the OLD INT value. If not, a step 65C
will reduce the INT value to zero. Otherwise, the subroutine proceeds from the inquiry 649 to an in~egrating step 651 where the decremental value is subtracted from the OLD INT valu~. The resul~ing difference is the previously described var~able value INT which now decreases at an average rate that varies with the size of the decrement. After either one of the steps 650 and 651 is executed, ~he next step 652 reloads the INT value ~in binary form~ into the old in~egrator value register where it is saved as OLD INT ~or the next pass through the FIG. 12 subroutine, and in a f~nal step 653 before returnin~ to the FIG. 10 routlne, the same value is loaded ~nto the aforesaid wheelslip correcti~n value re~tster.

: ` :

, ,; , ,:
. , -i7- 20~C-1532 So long as there are negative answers at bo~h of the inquiry steps 641 and 642 and the desired decrement is less than OLD INT, the integrating step fi51 will be executed every pass through the FIG. 12 subroutine. Consequently the OLD INT value is periodic~lly decremented 05 or reduced to a less positive value INT tha~ is saved as OLD INT for the next pass through, and ~his value becomes the new wheelslip correction value at the step 653. The size of each decrement, as determined by the step 647, is proportional to the ~moun~ by which the reference speed value exceeds K2, and the number of such decrements per seoond is constant (e.g.7 50). As a result, in the wheelslip recovery mode of operation the wheelslip correction value is reduced at a rate that varies with track speed, and the proportionality constant G is selected so that traction power will be restored at an optimum rate ~e.g.~ approximately 50 HP per second per powered axle). The step 648 will be operative when the reference speed is relatively low to prevent the decrement from being less than the minimum size D, thereby establishing a minimum rate of correction value decrease. D is selected so that when the reference speed value does not exceed K2 the wheelslip correction value will decrease at a prede~ermined constant rate ~corresponding to a desired rate of decrease of V), and now the rate of power restoration will vary with the track speed. As INT and the correction values approach zero, they will become smaller than the next decrement, whereupon they ~re returned to zero by the step 650, the answer at the inquiry step 641 will now change from no to yes, and the integrating step S51 becomes inoperative. Under most circumstances the answer to th~ ;nquiry step 6~2 will change from no to yes before the wheelslip correction value has been reduced to zer~ by the step 6~0 or 651, whereby the step 644 will be effective as previously explain~d to reset the integratin~ means so that the correction value and OLD INT
are promptly returned to zero.
After each pass through either the recovery/termination subroutine 618 ~FIG. 12) or the deration subrouginQ 61~ (FIG. 113, or whenever negative answers are ob~ained at all three of the inquiry steps 602, 604 and 606 tequivalent to the logic steps 87 and 92 in FIC. 5 haYing low output states, whereby both ~he derat~on mode and ~he recovery mode are inact1ve)~ the reference value reducing subrout~ne 616 of the wheelslip mode control routine (FIG. 101 is executed to end the latter .

. . ..
.. ~ ........ .

1 3~65~0 routine. The reference valu~ reducing subroutine 616 in effect implements the f1fth subsystem 500 of the wheelslip cnntrol function 65. A sin~plified version of i~s presently preferred embodiment is illustrated functionally in FI~. 13, and a flowchart o~ sof~ware 05 suitable for i~plementing these functions is shown in FIG. 14.
As is illustrat2d in FIG. 13, the wheelslip correction value (which is the outpu~ of ~he sum~ing means 90 in FIG. S) is supplied to a summing point 502 where it is subtracted from a value supplied by memory means 504. The input of the ~emory ~eans sa4 is coupled to a lowest-Yalue gate 506 via a normally closed switch 508 and an output line 5l0 of the gate. The switch 508 is open only when the logic step 88 (FIG. 5) is in its TRIP state, as is true throughout the deration and recovery modes of operation. The gate 506 ha~ ~hree inputs that are respectively coupled ~o the lines 51, 57 and 64 (see FIG. 3), and lS it selects the lowest of the three different reference values on these lines ~voltage llmit, current limit, and power, respect~vely) as the value on its output line 510. Normally the locomotive propulsion system is being controlled in a constant horsepower range, in which case the pow~r reference value on the line 64 will be lower than the other two reference values. The selected value is supplied to the input of the memory means via the switch 508. The memory means 504 is so arranged that its output value is approximately the same as the value on the line SlO when the switch 508 changes states from clos~d to open. In other words, during any interval that the logic step 88 is in a TRIP state (i.e, while either the deration or recovery mode of operat~on is activated), the switch 508 is open and the me~ory ~eans 504 is effective to remember and save whatever value was on the line SlO at the moment of time ~hen this state began (i.e., at the start of a deration mode) and tc huld such Yalue substantially constant at its output.
As is illustrated in FIG. 13, the wheelslip correct10n value is subtracted from the output value of the memory means 504 at the summing point 502. The resulting difference is suppl~ed via a switch 5l2, a line 66a, and another least-value gate 514 to ~he output ~ine 66 of the whPelslip control funct~on 65, and it is also supplied to the input of a source 516 labeled "RAMP UP." The switch 512 has two alternatiYe positions or states: normally ~t ~s ~n the sta~e shown in FIG. l3, . ; .
.. , ~

_~,9_ 20LC-1532 thereby connecting the line 66a to the output of the source 516; but whenever the logic step 88 is in a TRIP state, the switch 512 is in its second state so as to connect ~he line 66a directly to the output of the summing point S02. The value on ~he line 66a (herein referred to 05 as "WS Volts") will depend on the state of the switch 512. So long as th;s switch is in its second or TRIP state, WS Vol~s will be the same as the difference value at the output of the summing poin~ 502, which value will now be equal to the relatively constant value that the memory means 504 is saving minus the wheelslip correction value dertved from the su~ming means 90 (FIG. 5). Otherwise, WS Volts is determined by the source 516 which is so arranged as to increase this value to a predetermined maximum 1 imi~ when activated by the state change of a normally closed switch 518 in response to the termination of the TRIP
state (i.e., at the end of a recovery mode of operation).
The least-value gate 514 has two inputs, one coupled to the line 66a and the other coupled to the line 64. The gate 514 selects the s~aller of its two input values as the reference signal value on the output line 66. It will be apparent that when the switch 508 is opened and the switch 512 concurrently changes to its second state at the s start of a deration mode of operat~on, the gate 514 will beco~e effective in conJunction with the summing point 502 to reduce the re~erence signal value on the line 66 to a level equal ta WS Volts, and this reduction will usually continue until the subsequen~ recovery mode of operation causes the INT v~lue and consequently the wheelslip correction value to decrease to zero, thereby terminating the TRIP
state of the logic step 88. During the dera~ion and recovery modes, WS
Volts (and consequently the reference value on the line 66) equals the saved, constant value at the output of the me~ory means 504 minus the variable corr~ction value. In effect, the saved value is a ~snapshotN
of the reference value at the moment of time when the deration mode is in~tiated. As a result, the degree of reference value correct10n (i.e., the amount of tract10n power reduction or deration~ is des~rably referenced to the control system parameters that existed at the start of deration, and once the wheelslip condit~on is cured and ~he recovery mode is concluded, the reference value on the line 66 will again equal its snzpshot value. Thls provides a smooth and stable transition from .
, ,. ~

2 OLI: -1532 wheelslip control to normal control of the locomotive propulsion system.
To provide the aforesaid WS INCONTRL indication, bistable polarity detecting means 520 is coupled to the output of a summing point 522 05 where the power referenoe value on the line 64 is compared with the WS
Volts on the line 66a. So long as WS Yolts is lower ~han thc reference value on the line 64, the polarity detector 521D has a high output state to indicate that the wheelslip contrul function 65 is in control;
otherwise the polarity detector will be 1n ~s low output state. The detector 520 will change from high to low states in response to ~S
Volts being increased or ramped up to ~he level of the reference valu~
on line 64 when the source 516 is activated at the end of a recovery mode, as previously described, or in response to the power reference value on the line 64 decreasing below WS Volts at any time during the deration or recovery mode of operat~on. In the latter event, this state change de-activates the logic steps 91 and 88 (see FIG. 5), thereby terminating the TRIP state and at the same time c~using the reset function 322 to reset the integrator 310 in the subsystem 300 ~see FI6. 8), whereupon the INT value and consequently the wheelslip correction value are immediately re~urned to zero. The least-value gate 514 is atways e~fective to prevent the reference signal value on the output line 66 from exceeding the value on the line 64.
FIG. 14 is a flow chart of the presently preferred soft~are for implementing the reference signal reducing subroutine 616 shown functionally in FIG. 13. This subroutine is en~ered at an inquiry step 661 th~t tests the state of the aforesaid bistable trip flag. As previously explained ~see the deseription of step 608 in ~I~. 10 and of step 645 in FI~. 12), the tr~p flag will be in its true state (equivalent to the TRIP state of the OR logic step 88 in FI&. 5) during the derat~on and reeovery modes o~ operation. In this event the answer to the inquiry 161 is yes, and the next step ~62 is to ~est whether or not a bistable Nslip in processn flag is in a set or ~rue state. The latter step will have a nega~ve answer the ~irst time i~ ~s executed (i.e., on the first pass through the FIG. 14 subroutine af~er the trip flag has been changed to its true sta~e by the step 60~ [FIG. 101 at the beg~nning of a derat~on interval). Consequently the program will proceed fro~ th~s step to a leas~-value select~ng step ~63 that reads .
. - , : .

. . .
t ,:

,~

1 32~520 the lowest one of the reference values on the three lines 51, 57 and 64 and loads it into a temporary register (herein referred to as the "old wheelsl;p Yolts" register) where it is saved as "Old WS Volts." The ste~ 663 is followed by an initializing step 664 wherein the value 05 loaded in the old wheelslip vol~s register and the values saved in the aforesaid old integrator value register and wheelslip correction value register are respectively incremented by an amount equal to 1/Nth of Old WS Yolts, where N is a predeterminPd integer (e.g., 16). The incremented value in the old wheelslip volts register is now saved or remembered as "V-Sav" which is approximately the same as the lowest reference value at this time.
hfter the step 654 is execu~ed, the wheelsl;p correction value derived at either the last step 638 in FIG. 11 (deration mode~ or the last step 653 in FIG. 12 ~recovery mode) is subtracted from the V-Sav value at a step 665 ~equivalent to the summing point 502 in FIG. 13), and the difference is saved as WS Volts. The next step 666 tests the slip in process flag again. Initially the state of this flag is not true, and a step 667 is operative to change it to true before executing the succeeding inquiry step 668. Once the slip in process flag is thus set in its true state, during each subsequent pass through the FI6. 14 subroutine while the trip flag is true the program will proceed directly from the step 662 to the step 665, thereby bypassing the steps 663 and 664, and it will proceed directly from the step 666 to the step 668, thereby bypassing the step 667. As a result, the V-Sav value is loaded in the old wheelslip volts register at step 664 only during the first pass through this subroutine, and this value is not updated or changed again throughout the ensuing interval that the trip flag remains in ;ts true state.
In the step 668 of the FIG. 14 subroutine, WS Volts is compared with the power reference value on the line 64. So long as the former is lower than the latter, the subroutine proceeds from the step 668 to a step 670 where the setting of the aforesaid "wheelslip in control"
flag is changPd, if this flag were then in its false or low state9 to its true or high state (WS INCONTRL), and the FIG. 14 subrout~ne will then return to the wheelsltp mode control routine (FI6. lD~ via a f~nal step 671 that makes the reference value on the output line 66 the same as WS Volts. Otherwise the subroutine would proceed from the step 668 . . , , , ' ' , to an alternatlve step 672 where the same flag is reset, if necessary, to its false state, and the final step 673 will now make the reference value on the QUtpUt line 66 equal ~o the power reference value on the line 64. tIt will be apparent that the steps 668, 670 and 672 perform o5 the same functions as the summing point 522 and the bistable polarity detector 520 in FI6. 13, and the steps 668, 671 and 673 perform the same function as the least-value gate 514.) ~ henever the answer to the inquiry 668 chanses from yes to no, the step 672 resets the wheelslip in control flag which eauses the recovery mode of operatton to terminate. In practice this will usua71y occur before the OLD INT value (and hence the wheelslip correction value) is reduced to zero, in which event the answer to the inquiry step 641 in the recovery/termination subroutine (FIG. 8) is no, the answer to the inquiry step 642 is yes, the steps 644 and 645 are exeeuted to terminate the recovery mode, and there is an immediate, smooth, and stable transition from wheelslip control to the normal motoring mode of operation of the locomotive propulsion control system without any discontinuity in the power reference value on the output line 66. Note that this response would be obtained even if the referencD value on the input line 64 were to increase by as much as six pereen~ during the wheelslip control mode (as could happen if track speed were increasing3. As was previously explained (see the description of the initializing step 664), throughou~ the wheelslip control ~ode the value stored in the o1d wheelslip volts register (V-Sav) preferably is slightly higher than (e.g., 106%) the value of the reference signal that existed on the line 6~ at the mo~ent of time when the deration interval was initiated (i.e., when the state of ~he trtp flag changed from false to true~, and therefore ~S Volts will be just equal tq this higher value whenever the wheelslip correc~ion value is reduced to zero. In other words, the power re~erence value could be slightly higher at the conclusion of recovery than a~ the start of deration if required to r~store the sam~ eonstant level of tractlon power in spite of increased track speed.
After the trip flag ls returned to ~s reset or false state at the end of a recovery interval (see step 6~5 ~n FIG. 12), the answer to the in~tial inqu~ry s~ep 661 of ~he FI~. 14 subrout~ne w~l~ be no. In thls event, the step 6hl is followed by a step 675 wh1ch ensures that the . . ~ , . . .

1 ~26520 -52.1- 20LC--1532 slip in process flag is in its reset or false state. Then the saved value of WS Volts is checked to determine whether or not it is less than the aforesaid maximum limit "VLIM" which is greater than the highest reference value expected on the line 64. If not, the program 05 would proceed directly from step 676 to step 668. Otherwise, it proceeds to the step 66R via an additional 677 that increments WS Volts by a predetermined amount VLIM/X. The denominator X is selected so that WS Volts will increase to its maximum limit at a desirably fast rate (corresponding, for example, to the rectified output voltage of the main alternatnr 12 increasing at approximately 16 volts per second). It will now be appreciated that tha steps 676 and 677 perform the same function as the ramp-up source 516 shown in FIG. 13, and they are operative whenever the trip flag is not in its true s~ate to ensure a negative answer at the inquiry step 668.
While a preferred embodiment of the invention has been shown and described by way of illustration, various modifications thereof will probably occur to persons skilled in the art. It is therefore intended by the concluding claims to cover all such changes and modifications as fall within the true spirit and scope of this invention.

, :, ,

Claims (91)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:
1. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest ant lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;
c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
e. multiplying means for deriving a voltage error value related to said speed error value by a variable gain that is a predetermined function of said reference speed value;
f. integrating means operative when the actual value of said one speed differs from the desired value in a predetermined direction for obtaining a wheelslip correction value that increases at an average rate that varies with said voltage error value; and g. means associated with said integrating means for reducing the value of said reference signal by an amount corresponding to said correction value.
2. The wheelslip control means of claim 1 wherein said variable gain increases from a predetermined minimum limit as said reference speed value increases from zero, whereby the ratio of said voltage error value to said speed error value is dependent on said reference speed value.
3. The wheelslip control means of claim 2 wherein said gain varies non-linearly with said reference speed value.
4. The wheelslip control means of claim 1 wherein said multiplying means includes means for setting said variable gain at a value that varies, between predetermined minimum and maximum limits, in accordance with approximately the second power of said reference speed value.
5. The wheelslip control means of claim 1 wherein said third means is operative to vary said normally desired maximum difference as a predetermined function of said reference speed.
6. The wheelslip control means of claim 5 wherein said third means includes means for preventing said normally desired maximum difference from decreasing below a predetermined minimum speed.
7. The wheelslip control means of claim 6 wherein said third means includes means for setting said maximum difference at a speed that is a variable percentage of said reference speed, said percentage varying inversely with said reference speed as said reference speed varies between predetermined first and second values.
8. The wheelslip control means of claim 7 wherein said percentage is variable between predetermined maximum and minimum limits and is not less than said minimum limit when said reference speed is higher than said second value.
9. The wheelslip control means of claim 1 in which said predetermined one speed is the highest speed detected by said first means, said summing means includes means for comparing the lowest speed detected by said first means and the reference speed indicated by said second means, and said summing means is so arranged that said speed error value varies with the amount, if any, that said highest speed exceeds the sum of said maximum difference and either said reference speed of said lowest speed, whichever is lower.
10. The wheelslip control means of claim 1 in which said integrating means when operative periodically increments and saves said wheelslip correction value, the size of each increment being proportional to said voltage error value and the number of such increments per second being constant.
11. The wheelslip control means of claim 10 in which said integrating means includes means for preventing the wheelslip correction increments from decreasing below a predetermined minimum size.
12. The wheelslip control means of claim 11 wherein said last-mentioned means is operative if said reference speed exceeds a predetermined, relatively low value to set said minimum size at a predetermined first value and is operative otherwise to set said minimum size at a predetermined second value lower than said first value.
13. The wheelslip control means of claim 10 in which means is provided for returning said wheelslip correction value to zero when said integrating means is not operative.
14. The wheelslip control means of claim 10 wherein said reference signal reducing means includes memory means effective while said wheelslip correction value is greater than zero for saving a value approximately the same as the value of said reference signal at the time when said integrating means becomes operative, and wherein said reference signal reducing means is arranged when said memory means is effective to reduce said reference signal value to a level equal to said saved value minus said correction value regardless of the value of said command signal.
15. The wheelslip control means of claim 1 which further comprises additional means responsive to said speed error value and operative when said integrating means is operative and said speed error value exceeds a predetermined level to effect a step increase of said wheelslip correction value.
16. The wheelslip control means of claim 15 wherein the size of said step increase depends on the amount by which said speed error value exceeds said predetermined level up to a predetermined maximum limit.
17. The wheelslip control means of claim 15 wherein the size of said step increase is equal to the product of said voltage error value and a variable multiplier that increases from zero as said speed error value increases from said predetermined level.
18. The wheelslip control means of claim 1 which further comprises wheelslip detection means for detecting the number of separately driven wheels of the vehicle that are slipping, as indicated by excessively high second derivatives of the rotational speeds of the individual wheels, and in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if the number of slipping wheels is above a predetermined limit.
19. The wheelslip control means of claim 1 which further comprises means for respectively deriving the rates of change of rotational speeds of the separately driven wheels of the vehicle, means for comparing said rates of change and for detecting the highest one, and bistable means operative from a reset state to a set state in response to the highest rate increasing to at least a predetermined pickup level, and in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a wheel that is not slipping but will not change appreciably if said bistable means is in its set state.
20. The wheelslip control means of claim 19 wherein said predetermined one speed is the highest speed detected by said first means, said speed error value is positive whenever the actual highest speed is greater than desired, said integrating means is operative when said speed error value is positive, and said second means includes means for reducing said reference speed value and for ensuring that said bistable means is in its reset state whenever the lowest speed detected by said first means is less than said reference speed.
21. The wheelslip control means of claim 1 which further comprises means for respectively deriving the rates of change of rotational speed of the separately driven wheels of the vehicle, means for comparing said rates of change and for detecting the lowest one, and bistable means operative from a reset state to a set state in response to the lowest rate increasing to at least a predetermined pickup level, and in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a wheel that is not slipping but will not change appreciably if said bistable means is in its set state.
22. The wheelslip control means of claim 21 wherein said predetermined one speed is the highest speed detected by said first means, said speed error value is positive whenever the actual highest speed is greater than desired, said integrating means is operative when said speed error value is positive, and said second means includes means for reducing said reference speed value and for ensuring that said bistable means is in its reset state whenever the lowest speed detected by said first means is less than said reference speed.
23. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;

c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
e. integrating means associated with said summing means and operative when the actual value of said one speed differs from the desired value in a predetermined direction for obtaining a wheelslip correction value that increases at an average rate that is a predetermined function of said speed error value;
f. additional means associated with said integrating means and operative when the actual value of said one speed does not differ from said desired value in said predetermined direction and said wheelslip correction value is greater than zero for decreasing said correction value at an average rate that is independent of said speed error value;
and g. means associated with said integrating means for reducing the value of said reference signal by an amount corresponding to said correction value.
24. The wheelslip control means of claim 23 wherein said additional means includes means responsive to said reference speed value for varying said rate of decrease as a predetermined function of said reference speed.
25. The wheelslip control means of claim 24 wherein said rate of decrease varies with said reference speed so long as said reference speed exceeds a predetermined value.
26. The wheelslip control means of claim 25 wherein said rate of decrease equals a predetermined minimum rate whenever said reference speed is less than said predetermined value.
27. The wheelslip control means of claim 24 in which said additional means is arranged when operative periodically to decrement and save said wheelslip correction value, the size of each decrement being dependent on said reference speed value and the number of such decrements per second being constant.
28. The wheelslip control means of claim 27 wherein the wheelslip correction decrements are not less than a predetermined minimum size so long as said correction value equals or exceeds said minimum size.
29. The wheelslip control means of claim 27 wherein said additional means is inoperative to decrement said wheelslip correction value when said correction value has returned to zero.
30. The wheelslip control means of claim 23 wherein said predetermined one speed is the highest speed detected by said first means, said speed error value is positive whenever the actual highest speed is greater than desired, said integrating means is operative when said speed error value is positive, said reference signal reducing means includes memory means effective so long as said wheelslip correction value is greater than zero for saving a value approximately the same as the value of said reference signal at the time when said speed error increases from zero to a positive value, and said reference signal reducing means is arranged when said memory means is effective to reduce said reference signal value to a level equal to said saved value minus said correction value regardless of the value of said command signal .
31. The wheelslip control means of claim 30 for a propulsion system having regulating means responsive to said command signal for normally determining the value of said reference signal, wherein said reference signal reducing means includes means for preventing the reference signal from exceeding the value normally determined by said regulating means.
32. The wheelslip control means of claim 31 which further comprises bistable means having first and second states, said bistable means being in its first state whenever the reference signal value normally determined by said regulating means is lower than said reduced level and being arranged upon changing from second to first states to reset said integrating means, whereupon said correction value returns to zero.
33. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;
c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
e. means responsive to said speed error value for deriving a voltage error value that is a predetermined function of said speed error value;
f. integrating means operative when the actual value of said one speed differs from the desired value in a predetermined direction for obtaining a wheelslip correction value that increases at an average rate that varies with said voltage error value;
g. memory means effective while said wheelslip correction value is greater than zero for saving a value approximately the same as the value of said reference signal at the time when said integrating means becomes operative, said saved value being substantially constant; and h. means associated with said integrating means and operative while said memory means is effective for reducing said reference signal value to a level equal to said saved value minus said correction value regardless of the value of said command signal.
34. The wheelslip control means of claim 33 which further comprises additional means responsive to said speed error value and operative when said integrating means is operative and said speed error value exceeds a predetermined first level to increase said wheelslip correction value by an amount that is related to said voltage error value by a variable coefficient, whereby said correction value varies with the sum of said last-mentioned amount and an amount that is representative of the time integral of said voltage error value.
35. The wheelslip control means of claim 34 wherein said additional means includes means for varying said coefficient between zero and a predetermined maximum number as said speed error value varies between said first level and a predetermined second level higher than said first level.
36. The wheelslip control means of claim 33 wherein said voltage error value deriving means is so arranged that said voltage error value is related to said speed error value by a gain that varies, between predetermined limits, with said reference speed value.
37. The wheelslip control means of claim 33 for a propulsion system having regulating means responsive to said command signal for normally determining the value of said reference signal, wherein said reference signal reducing means includes means for preventing the reference signal from exceeding the value normally determined by said regulating means.
38. The wheelslip control means of claim 37 which further comprises bistable means having first and second states, said bistable means being in its first state whenever the reference signal value normally determined by said regulating means is lower than said reduced level and being arranged upon changing from second to first states to reset said integrating means, whereby said correction value then returns to zero.
39. The wheelslip control means of claim 38 wherein said predetermined one speed is the highest speed detected by said first means, said speed error value is positive whenever the actual highest speed is greater than desired, and said integrating means is operative when said speed error value is positive.
40. The wheelslip control means of claim 39 for a propulsion system having means for sensing the magnitude of electric power output of the source, wherein said regulating means is operative to vary said reference signal value as necessary to minimize any error between the sensed magnitude of power and a desired magnitude thereof.
41. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means for respectively deriving the rates of change of rotational speeds of said separately driven wheels;
c. third means for comparing said rates of change and for detecting the highest one;
d. bistable means operative from a reset state to a set state in response to the highest rate increasing to at least a predetermined pickup level;
e. wheelslip detection means for detecting the number of said separately driven wheels that are slipping, as indicated by excessively high derivatives of the rates of change derived by said second means;

f. reference speed indicating means associated with said first means for providing a reference speed value that normally varies with the rotational speed of a vehicle wheel that is not slipping but that will not change appreciably if either the number of slipping wheels is above a predetermined limit or said bistable means is in its set state;
g. means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
h. summing means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
j. means responsive to said speed error value for deriving a voltage error value that is a predetermined function of said speed error value;
k. integrating means operative when the actual value of said one speed differs from the desired value in a predetermined direction for obtaining a wheelslip correction value that increases at an average rate that varies with said voltage error value; and m. means associated with said integrating means for reducing the value of said reference signal by an amount corresponding to said correction value.
42. The wheelslip control means of claim 41 wherein said voltage error value deriving means is so arranged that said voltage error value is related to said speed error value by a variable gain that increases from a predetermined minimum limit as said reference speed value increases from zero.
43. The wheelslip control means of claim 41 wherein said reference speed indicating means is so arranged that normally said reference speed value will change at a rate corresponding to the rate at which said lowest speed is varying while the vehicle is being propelled by the motors.
44. The wheelslip control means of claim 43 wherein said reference speed indicating means includes means for preventing the rate of change of said reference speed value from exceeding a predetermined maximum rate.
45. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;
c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
e. wheelslip detection means for detecting the number of separately driven wheels that are slipping, as indicated by excessively high second derivatives of the rotational speeds of the individual wheels;
f. additional means coupled to said wheelslip detection means for determining whether or not the number of slipping wheels is above a predetermined limit;
g. means associated with said additional means and said summing means and effective when the number of slipping wheels is above said limit to change the desired value of said one speed by a variable amount that increases with the length of time that the number of slipping wheels is above said limit, said desired-value change tending to increase said speed error value in a positive direction;
h. means associated with said summing means and operative when said speed error value is positive for obtaining a wheelslip correction value that is a predetermined function of said speed error value; and J. means responsive to said wheelslip correction value for reducing the value of said reference signal by an amount corresponding to said correction value.
46. The wheelslip control means of claim 45 wherein said predetermined one speed is the highest speed detected by said first means, said speed error value is positive whenever the actual highest speed is greater than desired, and said summing means is so arranged that said desired speed equals the sum of said reference speed plus said normally desired maximum difference minus said variable amount.
47. The wheelslip control means of claim 45 wherein said reference signal reducing means includes memory means effective while said wheelslip correction value is greater than zero for saving a value approximately the same as the value of said reference signal at the time when said speed error value becomes positive, and wherein said reference signal reducing means is arranged when said memory means is effective to reduce said reference signal value to a level equal to said saved value minus said correction value regardless of said command signal.
48. The wheelslip control means of claim 45 wherein said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if the number of slipping wheels is above said predetermined limit.
49 The wheelslip control means of claim 48 wherein said second means is so arranged that said reference speed value will not change appreciably if the number of slipping wheels equals or exceeds said predetermined limit.
50. The wheelslip control means of claim 48 in which said means for obtaining said wheelslip correction value is operative when said speed error value is positive and exceeds a predetermined level to increase said wheelslip correction value by an amount that varies with the difference between said speed error value and said predetermined level.
51. The wheelslip control means of claim 45 wherein said variable amount increases from zero at a rate that varies with the maximum difference speed determined by said third means.
52. The wheelslip control means of claim 51 wherein said desired-value changing means includes timing means for measuring the time that the number of slipping wheels is above said limit, said variable amount being proportional to the product of the measured time and said maximum difference.
53. The wheelslip control means of claim 51 in which said variable amount is proportional to the product of said maximum difference speed and a variable factor that increases from 0 toward a maximum of 1.0 with the length of time that the number of slipping wheels is above said limit, and in which said desired-value changing means includes means operative in response to the number of slipping wheels decreasing from above said limit to a number not above said limit for preventing said variable factor from decreasing to 0 until the number of slipping wheels decreases below said limit, whereupon said last-mentioned means resets said factor to 0.
54. The wheelslip control means of claim 51 wherein said desired-value changing means includes means operative when said maximum difference is less than a predetermined minimum speed to prevent said rate of increase from falling below a predetermined minimum rate.
55. The wheelslip control means of claim 51 wherein said variable amount increases at a variable rate from zero to an upper limit proportional to the maximum difference determined by said third means.
56. The wheelslip control means of claim 55 wherein said upper limit is at least approximately twice said maximum difference.
57. The wheelslip control means of claim 45 in which said desired-value changing means includes means for abruptly reducing said variable amount to zero in response to the number of slipping wheels decreasing from above to below said predetermined limit.
58. The wheelslip control means of claim 57 in which said desired-value changing means is normally effective to hold said variable amount relatively constant so long as the number of slipping wheels equals said predetermined limit.
59. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;
c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;

e. means for respectively deriving the rates of change of rotational speeds of the separately driven wheels of the vehicle, for comparing said rates, and for detecting the lowest one;
f. bistable means operative from a reset state to a set state in response to the lowest rate increasing to at least a predetermined pickup level;
g. means associated with said summing means and effective when said bistable means is in its set state to change the desired value of said one speed by a variable amount that increases with the length of time that said bistable means is in its set state, said desired-value change tending to increase said speed error value in a positive direction;
h. means associated with said summing means and operative when said speed error value is positive for obtaining a wheelslip correction value that is a predetermined function of said speed error value; and j. means responsive to said wheelslip correction value for reducing the value of said reference signal by an amount corresponding to said correction value.
60. The wheelslip control means of claim 59 wherein said means for obtaining said wheelslip correction value is operative when said speed error value is positive and exceeds a predetermined level to increase said wheelslip correction value by an amount that varies with the difference between said speed error value and said predetermined level.
61. The wheelslip control means of claim 59 wherein said variable amount increases from zero at a rate that varies with the maximum difference speed determined by said third means.
62. The wheelslip control means of claim 61 wherein said desired-value changing means includes timing means for measuring the time that said bistable means is in its set state, said variable amount being proportional to the product of the measured time and said maximum difference.
63. The wheelslip control means of claim 61 wherein said desired-value changing means includes means operative when said maximum difference is less than a predetermined minimum speed to prevent said rate of increase from falling below a predetermined minimum rate.
64. The wheelslip control means of claim 61 wherein said variable amount increases at a variable rate from zero to an upper limit proportional to the maximum difference determined by said third means.
65. The wheelslip control means of claim 64 wherein said desired-value changing means includes means operative when said maximum difference is less than a predetermined minimum speed to prevent the rate at which said variable amount increases from falling below a predetermined minimum rate.
56. The wheelslip control means of claim 64 wherein said upper limit is at least approximately twice said maximum difference.
67. The wheelslip control means of claim 59 in which said means for deriving, comparing and detecting the rates of change of wheel speeds is arranged to detect both the lowest and the highest rates of the separately driven wheels, in which second bistable means is provided, said second bistable means being operative from a reset state to a set state in response to the highest rate increasing to at least said predetermined pickup level, and in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if said second bistable means is in its set state.
68. The wheelslip control means of claim 59 wherein said reference signal reducing means includes memory means for saving a value that normally is approximately the same as said reference signal value, said memory means being effective so long as said speed error value is positive to hold said saved value substantially constant, and wherein said reference signal reducing means is arranged when said memory means is effective to reduce said reference signal value to a level equal to said saved value minus said correction value regardless of said command signal.
69. The wheelslip control means of claim 59 which further comprises wheelslip detection means for detecting the number of separately driven wheels of the vehicle that are slipping, as indicated by excessively high derivatives of the rates of change of rotational speeds of the individual wheels, and additional means coupled to said wheelslip detection means for determining whether or not the number of slipping wheels is above a predetermined limit, and in which said desired-value changing means is effective when either said bistable means is in its set state or the number of slipping wheels is above said limit.
70. The wheelslip control means of claim 69 in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if either said bistable means is in its set state or the number of slipping wheels is above said limit.
71. The wheelslip control means of claim 69 wherein said variable amount increases from zero at a rate that varies with the maximum difference speed determined by said third means.
72. The wheelslip control means of claim 71 in which said variable amount is proportional to the product of said maximum difference speed and a variable factor that increases from 0 toward a maximum of 1.0 with the length of time that said bistable means is in its set state or the number of slipping wheels is above said limit, and in which said desired-value changing means includes means operative while said bistable means is in its reset state and the number of slipping wheels is not above said limit for preventing said variable factor from decreasing to 0 until the number of slipping wheels decreases below said limit, whereupon said factor is reset to 0.
73. The wheelslip control means of claim 69 wherein said means for obtaining said wheelslip correction value includes: proportional means operative when said speed error value is positive and exceeds a predetermined level for deriving a variable value that depends on the amount by which said speed error value exceeds said predetermined level; and means for increasing said correction value by an amount equal to said variable value.
74. The wheelslip control means of claim 73 wherein said means for obtaining said wheelslip correction value includes integrating means operative when said speed error value is positive for deriving a second variable value that increases at a an average rate that depends on said speed error value, said wheelslip correction value comprising the sum of said variable values.
75. The wheelslip control means of claim 59 for a propulsion system comprising means for deriving a second feedback signal having a value representative of the magnitude of current in said plurality of motors, which further comprises means for varying said pickup level as a function of the value of said second feedback signal.
76. The wheelslip control means of claim 75 wherein said pickup level varying means includes means for determining maximum and minimum limits between which said pickup level can vary, said pickup level remaining equal to said minimum limit when said second feedback value is below a first value and remaining equal to said maximum limit when said second feedback value is relatively high.
77. The wheelslip control means of claim 59 in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if said bistable means is in its set state.
78. The wheelslip control means of claim 77 in which said means for deriving, comparing and detecting the rates of change of wheel speeds is arranged to detect both the lowest and the highest rates of the separately driven wheels, in which second bistable means is provided, said second bistable means being operative from a reset state to a set state in response to the highest rate increasing to at least said predetermined pickup level, and in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if either one of said bistable means is in its set state.
79. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;
c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. fourth means for respectively deriving the rates of change of rotational speeds of the separately driven wheels of the vehicle;
e. fifth means associated with said fourth means for comparing said rates of change and for providing a rate value representative of the highest rate of change;
f. bistable means operative from a reset state to a set state in response to said rate value increasing to at least a predetermined pickup level;
g. first summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
h. integrating means associated with said first summing means and operative when the actual value of said one speed differs from the desired value in a predetermined direction for obtaining a first variable value that increases at an average rate that is a predetermined function of said speed error value;
j. additional means effective when said bistable means is in its set state for obtaining a second variable value that depends on the value of said feedback signal;
k. second summing means associated with said integrating means and said additional means for deriving a wheelslip correction value that varies with the sum of said first and second variable values; and m. means responsive to said wheelslip correction value for reducing the value of said reference signal by an amount corresponding to said correction value.
80. For inclusion in a propulsion system of a traction vehicle having a plurality of wheels that are subject to slipping with respect to the surface on which the vehicle travels, the system comprising a plurality of adjustable speed electric motors mechanically coupled in driving relationship to separate wheels of the vehicle, a controllable source of electric power the output of which is electrically coupled in energizing relationship to the respective motors, means associated with the source of power for varying the magnitude of its output current or voltage in accordance with a variable control signal, means for deriving a feedback signal representative of the actual current or voltage magnitude, and a plurality of means for respectively sensing the rotational speeds of the separately driven wheels, said control signal being provided by a controller to which the speed sensing means are coupled and which is operative to vary the control signal as necessary to minimize any difference between the feedback signal and a reference signal the value of which normally depends on the value of a variable command signal, improved wheelslip control means comprising:
a. first means for comparing the rotational speeds of the separately driven wheels of the vehicle and for detecting the highest and lowest speeds, respectively;
b. second means associated with said first means for providing a reference speed value indicative of the rotational speed of a vehicle wheel that is not slipping;
c. third means for determining a normally desired maximum difference between the reference speed indicated by said second means and a predetermined one of the speeds detected by said first means;
d. fourth means for respectively deriving the rates of change of rotational speeds of the separately driven wheels of the vehicle;
e. fifth means associated with said fourth means for comparing said rates of change and for providing a rate value representative of the highest rate of change;
f. bistable means operative from a reset state to a set state in response to said rate value increasing to at least a predetermined pickup level;

g. first summing means associated with said first, second and third means for deriving a speed error value representative of the algebraic sum of said predetermined one speed, said reference speed and said maximum difference, said speed error value being zero whenever said one speed has a desired value;
h. integrating means associated with said first summing means and operative when the actual value of said one speed differs from the desired value in a predetermined direction for obtaining a first variable value that increases at an average rate that is a predetermined function of said speed error value;
j. additional means associated with said fifth means and effective when said bistable means is in its set state for obtaining a second variable value that depends on said rate value;
k. second summing means associated with said integrating means and said additional means for deriving a wheelslip correction value that varies with the sum of said first and second variable values; and m. means responsive to said wheelslip correction value for reducing the value of said reference signal by an amount corresponding to said correction value.
81. The wheelslip control means of claim 80 in which said second means is so arranged that said reference speed value normally varies with the rotational speed of a vehicle wheel that is not slipping but will not change appreciably if said bistable means is in its set state.
82. The wheelslip control means of claim 80 wherein said additional means is so arranged that said second variable value varies with the product of the rate value and the value of said feedback signal.
83. The wheelslip control means of claim 82 wherein said second variable value is a predetermined percentage of said product, and wherein means responsive to said reference speed is provided for determining said percentage.
84. The wheelslip control means of claim 82 for a propulsion system comprising means for deriving a second feedback signal having a value representative of the magnitude of current in said plurality of motors, which further comprises means for varying said pickup level as a predetermined function of the value of said second feedback signal.
85. The wheelslip control means of claim 80 which further comprises means responsive to said speed error value for deriving a voltage error value that is a predetermined function of said speed error value, and in which said integrating means is so arranged that said first variable value is representative of the time integral of said voltage error value.
86. The wheelslip control means of claim 85 which further comprises means associated with said first summing means and said voltage error value deriving means and operative when said speed error value exceeds a predetermined level for obtaining a third variable value related to said voltage error value by a coefficient that is a predetermined function of said speed error value, said second summing means being so arranged that said wheelslip correction value corresponds to the sum of said first, second and third variable values.
87. The wheelslip control means of claim 80 wherein said reference signal reducing means includes memory means effective while said wheelslip correction value is greater than zero for saving a value approximately the same as the value of said reference signal at the time when said correction value is first derived, said saved value being substantially constant, and wherein said reference signal reducing means is arranged when said memory means is effective to reduce said reference signal value to a level equal to said saved value minus said correction value regardless of the value of said command signal.
88. The wheelslip control means of claim 80 for a propulsion system comprising means for deriving a second feedback signal having a value representative of the magnitude of current in said plurality of motors, which further comprises means for varying said pickup level as a predetermined function of the value of said second feedback signal.
89. The wheelslip control means of claim 88 wherein said bistable means is operative from its set state to its reset state in response to said rate value decreasing to less than a predetermined fraction of said pickup level.
90. The wheelslip control means of claim 88 wherein said pickup level varying means is so arranged as to vary said pickup level with the adhesion level of the vehicle wheels on said surface, as indicated by the relative magnitude of motor current, as said adhesion level varies between predetermined first and second percentages.
91. The wheelslip control means of claim 90 wherein said pickup level varying means includes means for determining maximum and minimum limits between which said pickup level can vary, said pickup level remaining equal to said minimum limit when said adhesion level is lower than said first percentage and remaining equal to said maximum limit when said adhesion level is higher than said second percentage.
CA000607117A 1988-10-31 1989-07-31 Locomotive wheelslip control system Expired - Lifetime CA1326520C (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US07/265,377 US4896090A (en) 1988-10-31 1988-10-31 Locomotive wheelslip control system
US265,377 1988-10-31

Publications (1)

Publication Number Publication Date
CA1326520C true CA1326520C (en) 1994-01-25

Family

ID=23010184

Family Applications (1)

Application Number Title Priority Date Filing Date
CA000607117A Expired - Lifetime CA1326520C (en) 1988-10-31 1989-07-31 Locomotive wheelslip control system

Country Status (8)

Country Link
US (1) US4896090A (en)
CN (1) CN1020338C (en)
AU (1) AU619287B2 (en)
BR (1) BR8905551A (en)
CA (1) CA1326520C (en)
IN (1) IN171943B (en)
MX (1) MX164328B (en)
NZ (1) NZ229395A (en)

Families Citing this family (73)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
SE464860B (en) * 1989-10-24 1991-06-24 Asea Brown Boveri DRIVING SYSTEM FOR RAEL VEHICLES WITH LOAD SHARING BETWEEN DRIVER ENGINES
US5289093A (en) * 1989-11-08 1994-02-22 Gec Alsthom Sa Antispin and antilock methods for an electric traction vehicle
FI88015C (en) * 1990-12-17 1993-03-25 Kone Oy Procedure for generating speed setpoint for a lifting motor
US5428538A (en) * 1991-08-12 1995-06-27 Westinghouse Air Brake Company Sanding control system for railway vehicles
DE4135691C2 (en) * 1991-10-25 1998-04-16 Aeg Westinghouse Transport Arrangement for driving and braking control of vehicles that are equipped with several single wheel drive and braking modules
US5280223A (en) * 1992-03-31 1994-01-18 General Electric Company Control system for an electrically propelled traction vehicle
US5480220A (en) * 1992-08-26 1996-01-02 General Electric Company Method for inhibiting wheel slip in an electric alternating current induction motor powered vehicle
JP3387542B2 (en) * 1993-02-08 2003-03-17 株式会社ナブコ Load setting device for railway vehicles
US5436538A (en) * 1994-07-05 1995-07-25 Woodward Governor Company Locomotive wheel slip controller
US5740043A (en) * 1996-03-04 1998-04-14 Westinghouse Air Brake Company Relative spin speed traction control
US5841254A (en) * 1997-07-09 1998-11-24 General Electric Company Torque maximization and vibration control for AC locomotives
US5939846A (en) * 1997-09-04 1999-08-17 General Electric Company AC motorized wheel control system
US6012011A (en) * 1997-09-11 2000-01-04 Johnson; Chipley H. Traction control system and a method for remedying wheel-slippage
US6148269A (en) * 1998-07-20 2000-11-14 General Electric Company Wheel diameter calibration system for vehicle slip/slide control
US6028402A (en) * 1999-01-25 2000-02-22 General Electric Company Automatic rail characterization for adhesion system evaluation for AC locomotives
US5990648A (en) * 1999-01-29 1999-11-23 General Electric Company Method for detecting locked-axle conditions without a speed sensor
US6163121A (en) * 1999-01-29 2000-12-19 General Electric Company Torque maximization and vibration control for AC locomotives
US6104148A (en) * 1999-04-15 2000-08-15 General Electric Company System and method for controlling an AC traction motor without sensing motor rotation speed
US6194850B1 (en) 1999-08-23 2001-02-27 General Electric Company System and method for determining true ground speed in a locomotive
SE517013C2 (en) * 1999-10-12 2002-04-02 Abb Ab Apparatus and method for estimating the speed of a slip-ring asynchronous machine and its use
US6208097B1 (en) * 1999-12-06 2001-03-27 General Electric Company Traction vehicle adhesion control system without ground speed measurement
US6758087B2 (en) * 2000-04-28 2004-07-06 General Electric Company Method, system and storage medium for determining a vehicle reference speed
US6675771B2 (en) * 2001-12-28 2004-01-13 Visteon Global Technologies, Inc. Repeatability in control systems that utilize discretized feedback
US6812656B2 (en) * 2002-02-27 2004-11-02 Railpower Technologies Corp. Sequenced pulse width modulation method and apparatus for controlling and powering a plurality of direct current motors
US6813583B2 (en) * 2002-12-20 2004-11-02 General Electric Company Wheelspeed estimation system and method
JP4120504B2 (en) * 2003-07-30 2008-07-16 トヨタ自動車株式会社 Vehicle and vehicle control method
US7027900B2 (en) * 2003-09-19 2006-04-11 General Electric Company Enhanced locomotive adhesion control
KR100517110B1 (en) * 2003-10-14 2005-09-27 한국철도기술연구원 Apparatus for measuring simultaneously vibration and rotation speed of hauling motor and method thereof
US7084602B2 (en) * 2004-02-17 2006-08-01 Railpower Technologies Corp. Predicting wheel slip and skid in a locomotive
WO2005084335A2 (en) * 2004-03-01 2005-09-15 Railpower Technologies Corp. Cabless hybrid locomotive
WO2005097573A2 (en) 2004-03-30 2005-10-20 Railpower Technologies Corp. Emission management for a hybrid locomotive
WO2005114811A2 (en) * 2004-05-17 2005-12-01 Railpower Technologies Corp. Design of a large battery pack for a hybrid locomotive
CA2576871A1 (en) * 2004-08-09 2006-02-23 Railpower Technologies Corp. Regenerative braking methods for a hybrid locomotive
CA2576856C (en) 2004-08-09 2014-02-04 Railpower Technologies Corp. Locomotive power train architecture
CA2579174C (en) * 2004-09-03 2015-11-24 Railpower Technologies Corp. Multiple engine locomotive configuration
DE102005015658A1 (en) * 2005-04-06 2007-01-11 Bayerische Motoren Werke Ag Switching device for linking different electrical voltage levels in a motor vehicle
US7906862B2 (en) * 2005-04-25 2011-03-15 Railpower, Llc Multiple prime power source locomotive control
US7548032B2 (en) * 2005-08-23 2009-06-16 General Electric Company Locomotive speed determination
US20080264291A1 (en) * 2005-10-19 2008-10-30 Rail Power Technologies Corp Design of a Large Low Maintenance Battery Pack for a Hybrid Locomotive
US7778747B2 (en) * 2006-08-31 2010-08-17 National Railway Equipment Co. Adhesion control system for off-highway vehicle
DE102007021675A1 (en) * 2007-05-09 2009-01-15 Volkswagen Ag Method for controlling the optical axis of a vehicle headlight
US20080288132A1 (en) 2007-05-16 2008-11-20 General Electric Company Method of operating vehicle and associated system
ATE551222T1 (en) * 2007-06-27 2012-04-15 Mitsubishi Electric Corp CONTROL FOR ELECTRIC VEHICLES
US8280568B2 (en) * 2007-09-18 2012-10-02 Mitsubishi Electric Corporation Electric train car controlling device
EA023068B1 (en) * 2007-10-12 2016-04-29 Дженерал Электрик Компани System for dynamically affecting a force applied to a rail track through a rail vehicle axle
DE102008040366A1 (en) * 2008-07-11 2010-01-14 Robert Bosch Gmbh System e.g. antilock brake system, for controlling functional components of e.g. sports vehicle, has RAM connected with electrically EPROM utilized for storing control parameters, which parameterize system for different vehicle variants
US8310176B2 (en) * 2008-12-22 2012-11-13 Progress Rail Services Corp Traction control for DC electric motor
US8850990B2 (en) 2010-11-24 2014-10-07 Ztr Control Systems Method and apparatus for controlling and enhancing tractive effort in DC traction motors on locomotives
WO2012135258A2 (en) * 2011-03-29 2012-10-04 Glacier Bay, Inc. Generator
US8604728B2 (en) * 2011-05-11 2013-12-10 Trac Rail Inc. Method and apparatus for controlling dynamic braking on locomotives
US9209736B2 (en) * 2011-10-03 2015-12-08 General Electric Company System and method for traction motor control
KR101727329B1 (en) * 2011-10-19 2017-04-17 엘에스산전 주식회사 An apparatus and method for mesuring velocity of train
DE102011119207A1 (en) * 2011-11-23 2013-05-23 Audi Ag Method for controlling the operation of an arrangement of at least two electric machines and motor vehicle
FI124062B (en) 2012-10-03 2014-02-28 Konecranes Oyj Brake Control
CN103010049B (en) * 2012-12-13 2015-04-15 中国北车集团大连机车车辆有限公司 Anti-skidding control method for direct current transmission locomotive
CN103101541B (en) * 2012-12-13 2015-05-13 中国北车集团大连机车车辆有限公司 Anti-slip control method of direct-current drive locomotive
DE102013210361A1 (en) * 2013-06-04 2014-12-04 Siemens Aktiengesellschaft Method for determining at least one speed in a rail vehicle
US8831853B1 (en) * 2013-07-12 2014-09-09 Andrew Wyllie Barrowman Slip or loss of traction detector using the third derivative of rotational speed difference
CN104391177B (en) * 2014-11-17 2017-10-31 中车青岛四方机车车辆股份有限公司 EMUs side harmonics test system and method
US20160229433A1 (en) * 2015-02-06 2016-08-11 Electro-Motive Diesel, Inc. Automatic disabling of unpowered locked wheel fault detection
DE102015214176A1 (en) * 2015-07-27 2017-02-02 Continental Teves Ag & Co. Ohg Method for driver assistance with water smoothness on a road surface
JP6454251B2 (en) * 2015-10-06 2019-01-16 公益財団法人鉄道総合技術研究所 Derailment state detection device and derailment state detection method
CN105398455A (en) * 2015-11-17 2016-03-16 南车株洲电力机车有限公司 Double-power-source haulage motor and power control system thereof
AU2019207738B2 (en) * 2018-01-11 2022-06-02 Nsh Usa Corporation System for reprofiling a wheel set of a railway vehicle
CN108501768B (en) * 2018-03-29 2021-05-07 南京航空航天大学 Two-wheel speed control method based on Z-axis gyroscope and wheel speed difference
US11155135B2 (en) 2018-06-06 2021-10-26 Nordco Inc. Traction system for railcar movers
CN108791324B (en) * 2018-07-16 2023-06-20 湖南工业大学 Heavy-duty locomotive traction total amount consistency system and control distribution method
US10815644B2 (en) * 2018-09-19 2020-10-27 Deere & Company Tire size calibration and control system
EP3939820A4 (en) * 2019-03-15 2022-11-16 Hitachi, Ltd. Train control device and train control method
DE102021203898A1 (en) * 2021-04-20 2022-10-20 Siemens Mobility GmbH Sensor arrangement and method for determining the position of a rail vehicle
CN114795005A (en) * 2022-04-29 2022-07-29 添可智能科技有限公司 Cleaning equipment control method and cleaning equipment
EP4303088A1 (en) * 2022-07-04 2024-01-10 Dellner Bubenzer AB Controller for estimating individual axle weights of a rail vehicle, computer-implemented method therefor, computer program and non-volatile data carrier
CN116257069B (en) * 2023-05-16 2023-08-08 睿羿科技(长沙)有限公司 Unmanned vehicle formation decision and speed planning method

Family Cites Families (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3210630A (en) * 1963-03-15 1965-10-05 Gen Electric Plural motor drive with differential and synchronous slip control
US3437896A (en) * 1966-07-05 1969-04-08 Summit Electronics Inc Wheel slip detection system
US3541406A (en) * 1967-03-24 1970-11-17 Etablis Brissonneau Et Lotz Device for detecting and eliminating generalized wheel-slipping in electrically-propelled vehicles
US3482887A (en) * 1968-01-26 1969-12-09 Westinghouse Air Brake Co Combined anti-slip and anti-spin control for vehicle wheels
CH470282A (en) * 1968-04-17 1969-03-31 Oerlikon Maschf Rail locomotive with anti-skid device
SE342587B (en) * 1970-03-06 1972-02-14 Asea Ab
DE2127961C2 (en) * 1971-05-06 1983-10-27 BBC Aktiengesellschaft Brown, Boveri & Cie., 5401 Baden, Aargau Control device for the electromotive drive of rail vehicles
SE371615C (en) * 1973-04-06 1975-04-10 Asea Ab
US3982164A (en) * 1974-12-18 1976-09-21 General Motors Corporation Locomotive wheel slip control
US4075538A (en) * 1976-05-19 1978-02-21 General Electric Company Adaptive acceleration responsive system
DE2707047B2 (en) * 1977-02-18 1979-08-02 Siemens Ag, 1000 Berlin Und 8000 Muenchen Arrangement for detecting the skidding or sliding of the wheels of non-axled rail locomotives
JPS5517230A (en) * 1978-07-21 1980-02-06 Hitachi Ltd Control device for electric motor car
US4292572A (en) * 1979-07-26 1981-09-29 Dresser Industries, Inc. Method and apparatus for regulating electric motor armature currents
SU975467A1 (en) * 1980-12-12 1982-11-23 Московский Ордена Ленина И Ордена Трудового Красного Знамени Институт Инженеров Железнодорожного Транспорта Apparatus for detecting locomotive wheel skidding
CA1197597A (en) * 1982-03-11 1985-12-03 John A.I. Young Wheel slip control using differential signal
FR2523913A1 (en) * 1982-03-24 1983-09-30 Faiveley Ets METHOD FOR DETECTION AND CORRECTION OF WHEEL SKATING OF RAILWAY ENGINE ENGINE AND DEVICE FOR IMPLEMENTING SAME
ATE29118T1 (en) * 1983-09-28 1987-09-15 Bbc Brown Boveri & Cie SLIP LIMITING REGULATION FOR RAIL VEHICLES.
US4634887A (en) * 1985-06-24 1987-01-06 General Electric Company Load rate limiting means for a locomotive engine
DE3826075A1 (en) * 1988-07-30 1990-02-01 Hoechst Ag 1-OLEFIN ISO BLOCK POLYMER AND METHOD FOR THE PRODUCTION THEREOF

Also Published As

Publication number Publication date
AU619287B2 (en) 1992-01-23
CN1020338C (en) 1993-04-21
NZ229395A (en) 1991-11-26
MX164328B (en) 1992-08-03
AU3405089A (en) 1990-05-03
US4896090A (en) 1990-01-23
CN1042687A (en) 1990-06-06
IN171943B (en) 1993-02-13
BR8905551A (en) 1990-05-29

Similar Documents

Publication Publication Date Title
CA1326520C (en) Locomotive wheelslip control system
US4335337A (en) Control system for electric motor vehicle
US6208097B1 (en) Traction vehicle adhesion control system without ground speed measurement
US5992950A (en) Controlled stop function for locomotives
US5661378A (en) Tractive effort control method and system for recovery from a wheel slip condition in a diesel-electric traction vehicle
US4695941A (en) Loss of electrical feedback detector
CA1201166A (en) Locomotive governor control
US7467830B2 (en) Managing wheel slip in a locomotive
CN1031931C (en) Electric propulsion system with adaptive overspeed limit for traction vehicles
US3997822A (en) Method of controlling locomotive wheel slip
US5610819A (en) System for enhancing wheel traction in a locomotive by reapplication of excitation using an S-shaped curve
EP0145374A2 (en) Automatic speed control systems
HU218526B (en) Method and device for control of vehicles
US6758087B2 (en) Method, system and storage medium for determining a vehicle reference speed
JPH07500415A (en) Device for determining the critical driving torque of an automobile
EP0419578A1 (en) Locomotive differential wheel slip control.
US4463289A (en) Wheel slip control using differential signal
US4588932A (en) Slip-limiting control for rail vehicles
CN102205846A (en) Running speed adjustor for motor vehicle
EP0500906B1 (en) Locomotive dynamic brake control
KR910000771B1 (en) Running control method and running control system
KR102326629B1 (en) Control device and method for traction control for an electric drive system
CA1123496A (en) Locomotive wheel slip determination
EP0336046B1 (en) Control method of electric vehicle
JPH11103507A (en) Speed controller for car

Legal Events

Date Code Title Description
MKEX Expiry

Effective date: 20110125