Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS3642099 A
Publication typeGrant
Publication dateFeb 15, 1972
Filing dateAug 12, 1969
Priority dateAug 21, 1968
Publication numberUS 3642099 A, US 3642099A, US-A-3642099, US3642099 A, US3642099A
InventorsHirasawa Kotaro, Iwasaka Tatsuo, Kawatake Koichi, Matsuzawa Hideto, Yuminaka Takeo
Original AssigneeHitachi Ltd
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Group supervisory control system for elevators
US 3642099 A
Abstract
A group supervisory system for elevators includes a traffic demand detector for detecting traffic demands to provide a traffic demand signal. Also included is a pattern classifier for producing discriminant functions which are functions of the traffic demand signal and for determining an optimum traffic demand condition by selecting that discriminant function which represents a maximum value.
Images(16)
Previous page
Next page
Claims  available in
Description  (OCR text may contain errors)

United States Patent Yuminaka et al.

GROUP SUPERVISORY CONTROL SYSTEM FOR ELEVATORS Inventors: Takeo Yuminaka; Tatsuo lwasaka; llideto Matsuzawa, all of Katsuta-shi; Koichi Kawatake; Kotaro Hirasawa, both of Hitachi-shi, all of Japan Assignee: Hitachi, Ltd., Tokyo, Japan Filed: Aug. 12, 1969 Appl. No.: 849,441

Foreign Application Priority Data Feb. 15,1972

Primary Examiner-Bamard A. Gilheany Assistant Examiner-W. E. Duncanson, Jr. Attorney-Craig, Antonelli & Hill [57] ABSTRACT A group supervisory system for elevators includes a traffic de- Aug. 21, 1968 Japan ..43/59174 mand detector for detecting traffic demands to provide a traf- Dec. 27, 1968 Japan..... ....43/95477 fic demand signal. Also included is a pattern classifier for Jan. 31,1969 Japan ..44/6677 producing discriminant functions which are functions of the traffic demand signal and for determining an optimum traffic US. Cl. ..l87/29R demand condition by electing that discriminant function Int. Cl ...B66b U20 hich represents a maximum value, Field of Search 187/29 17 Claims, 22 Drawing Figures LUNCH TIME LUNCH CAGE LOAD AT DEPARTING CAGE TIME LUNCH K- ARRIVING CAGE --CLA$$|F|ER OTHERS -uP PEAK FD UP PDI E up CAGE DOWN HEAVY UP HEAVY k s02 DOWN P02 SECOND HEAVY DOWN o lNTERMlT' v PATTERN INTERMITTENT J UP CLASSIFIER UP. HEAVY UP DOWN ER NUMBER OF DOWN BALANCE a, FREE HALL CALLS UPPER ZONE BALANCE -1. wER

THIRD UPPER FREE R LOWER ZON PATTERN FLOOR CLASSIFIER LOWER FREE LG w il+| n:

H FIG. 60 II COMPARATOR A Vol DI I2 CMI II, T NTe COMPARATOR V; T I2 L V9 CV I1 COMPARATOR TC i I21 3/2 D3 Cha INVENTORS K0 ram: H RAsAwA ATTORNEYS mcminrmswn 3.642.099

sum user 16 WDu 3 ml- ADI,

E XI i i 2 E Di MAXIMUM SELECTOR I LMLD XL 5 ADR INVENTORS rmso YuMI-NAKA, rATSuo IWASAKAI HIDETO mArsuAAwA, KOICHI KAWATAKQ and ATTORNEY)- PAIENTEUFEB 1s i972 3, 642.099

sum as 0F 16 OFF 0 v O INPUT TAOYL COMPARATOR l 12 \CM| FIG. 7 (II TB COMPARATOR rz ww 12 W Va cv CM COMPARATOR FIPD WU PATTERN TU X A? A HD ACLASSIFIER BI FIG. 20 2RD PATTERN CLASSIFIER 12.

2AND

INVENTORS rAKEo Y M h rnrsuo IWASAKA,

K T'ARO HIRASA A ATTORNEYS PAIENTEBFEB 15 m2 SHEET 08 0F 16 arm :50 Enm i mmnm v v in 2mm 3 m-mwm m m oIcHJ; KAWA farm; and

Hum ro MRTSMLAWA' KOTAKo HIRASAWA ATTORNEYS PATENTEDFEB 1 5 m2 FIG.

WT DET MEANS FIG. Em

E m DEPARTING CAGE LOAD LUNCH TIME #5 x LUNCH TIIME [=3 OTHERS I5: OTHERS P? x LUNCH TIME i=8 INVENTORS TAKBO YLQMINAKA, T'HTSHO IWASAKA,

Hrosro MA'rsHzAwA, Koren;- KAWAI'AKE a d K0 TARO H RRSAWA l -j d ATTORNEY- PATENTEUFEB I 5 m2 SHEET 11 0F 16 E MTQ Pr 954 M96 wzrrmamwo INVENTORS IWASA KR,

rAkEo YMMINAKA, TA suo HLDET-O q AWA. KOICHI KAWA AKE dnJ m m 140 HIRASAWA Mao M 4 ATTORNI Y5 wi 4 3x226 PATENTEDFEBI 5 I972 r sum 13 or "16 HALL LOAD BFUPWARDLY MOVING CAGE R OF UP CALL Sy 0 I NUMBER OF UPPER .m 3 20 wzoN E20 6 $5232 $3 25: d zsoa .6 9x:

NUMBE Q o mwmznz A 1 FIG; l6" -HEAv.Y UP EOI'HEAVYUP Pg: AIHEAVY oowwpg. INTERMQTTENT Pl M HEAVY UP DOWN Q: BALANCE P;

A: BALANCE i=2, E1: BALANCE B3 FIG. l8

?BALANcE P;

-UPPER FREE FLOOR A H A: LOWER- FREE.

FLOOR ZONE CALLS INVENTORS rAKto yumzNAkA, 'rAl'Suo IWASAKA' nror, -.',.,,4rsuw,, 501B": x wqnuu-z Ker-Ago nznnsn'wa t n I'LL.

. -ATTORNEY5 PAIENTEMEB 15 m2 3. s42 099 SHEET 1 8 HF 16 PC-il 620%) FIG. 2|

DISCR FUNCTION G EN Gal R3" TD PRaz Rf 3 wD QAPMP MAX- TB R32l (Ch MUM BIAND PR3: T 933 Tu SEL- l 0P3 l ECT NOT I! H0 634 J U Rs4| F. TE C (32AND R35| I (F7655 TM DISCR FUNCTION GEN T'MER TAA'EO yuMI/vAKA, TATSMa EYE ATTORNEY) GROUP SUPERVISORY CONTROL SYSTEM FOR ELEVATORS This invention relates to improvements in or relating to elevator group supervisory control system for. efficiently operating a plurality of elevators, especially a group of elevators provided in a building in juxtaposition with each other in accordance with the traffic demand for the elevators.

Recently, more and more large-sized buildings have been constructed so that the numbers of elevators and fioors thereof have correspondingly increased. Thus, it has become essential to control those elevators as a group rather than individually in such an operating program as to be able to efficiently cope with the traffic demand, thereby increasing the carrying power.

The traffic demand for an elevator group differs between buildings but in the case of a typical ofiice building, it may be classified into several patterns. Such patterns are called traffic demand conditions (referred to simply as demand conditions hereinafter), which may include lunch time, heavy up," heavy down," intermittent," balance, heavy up-down and so forth. Lunch time condition occurs due to the fact that at lunch time, the number of passengers arriving at and leaving the lunch floor suddenly increases. Heavy up "condition occurs due to the fact that the number of passengers going from the lower dispatching landing to upper middle floors increases before the start of the office hours or at the end of the lunch time. This heavy up condition" is sometimes further classified into up peak condition where there are an increased number of passengers required up direction during the rush hours in the morning for example and heavy up condition other than such up peak.

Heavy down condition occurs due to the fact that passengers working on upper floors go down to the lower dispatching landing all at once at the end of the office hours or at the start of the lunch time. This heavy down condition is often classified into down peak condition where there are an increased number of passengers required down direction and other heavy down condition, other than such down peak, as in the case ofheavy up.

"Heavy up-down condition" occurs when the number of passengers temporarily increases during the day time, with a balance substantially maintained between the number of passengers required up direction and that of passengers required down direction. As the number of passengers further decreases, balance condition occurs which may also be classified into upper free floor or lower free floor" where I there is a relatively great demand for the upper floor zone or lower floor zone and other balance condition, other than such upper free floor and lower free floor. During a "upper free floor" condition, in general, the number of upper zone calls is much greater than the number of lower zone calls. However, during a lower free floor condition, the number of lower zone calls is much greater than the upper zone calls.

The foregoing demand conditions may be conceivable, but it is impossible to directly grasp them since the traffic demands for elevators include a variety of factors and always continuously change. However, the aforementioned classification of the demand conditions is possible by effecting analysis and synthesis based upon traffic demand signals X (referred to simply as demand signals hereinafter) consisting of several typical factors such as shown in FIG. 1. In an attempt to control the driving of the elevators, the control is effected with respect to each elevator group in accordance with an operating pattern suited to the thus determined demand conditions. This operating pattern is called operating program which is normally made to correspond to the respective demand condi tions as shown in FIG. 1. In case the demand condition corresponds to up peak condition for example, thena group of elevators are operated in accordance with the program of peak up operation."

However, if the operating program of heavy up-down operation" cannot be carried out in the case where the demand condition corresponds to heavy up-down," then heavy operation" may be instructed.

The group control system is so designed as to select that demand condition which corresponds to a particular time from the foregoing limited demand signals so as to efiiciently operate plural elevators. There have heretofore been proposed a variety of such systems. However, such conventional systems have such a disadvantage that the means for detecting demand signals and means for selecting a particular demand condition become large-sized and complicated as the number of elevators to be group-controlled and elevator floors increase, since all of them are constituted by a mere combination of contact type or contactless type relays.

It is an object of the present invention to provide a group supervisory control system for elevators which can be easily constituted by solid state devices (miniaturized), wherein the adjustment of piecewise boundary line (or piecewise boundary face) can be easily achieved.

Another object of the present invention is to provide a system whereby an l-number of variable parameters defined by l-parameter demand signals X consisting of several factors of traffic demand are sectioned in such a manner as to be most suitable to each demand condition, thereby making it possible to select a demand condition closer to the actual traffic demand for the elevators.

Still another object of the present invention is to provide a system wherein use is made of at least two sets of pattern classifiers, demand signals consisting of factors of traffic demand partly and overlapping each other or completely different from each other are supplied to the respective pattern classifiers, the number of parameters of each of the pattern classifiers is reduced to simplify the system, and the volume in which the demand signals are present is reduced so that an accurate faculty of discrimination can be achieved by setting up a small number of prototype points.

Still another object of the present invention is to provide a system wherein use is made of at least two sets of pattern classifiers, demand signals consisting of the same factors of traffic demand are supplied to the respective pattern classifiers, the discriminant functions of the lower rank pattern classifiers are set up to be greater than those of the higher rank pattern classifiers to make the resolutions of the former than those of the latter, thereby making it possible to achieve more accurate discrimination.

A further object of the present invention is to provide a traffic demand detecting means wherein a load in each elevator cage is sampled and held every time a particular floor is reached or left so that the traffic demand for the particular floor is continuously detected, and the demand for the floors between the particular floor and the terminal floor is neglected when the former is close to the latter, thereby accurately'detecting traffic demands.

A further object of the present invention is to provide a traffic demand detecting means which is capable of detecting the number of all hall calls for each floor zone and number of up" hall calls and down" hall calls at all the floors merely by using one contact (including a contactless type one) corresponding to each hall call.

A still further object of the present invention is to provide a maximum (or minimum) selecting means constituted by a combination of diodes and comparator means utilizing the forward voltages of the diodes.

Other objects, features and advantages of the present invention will become apparent from the following description taken in conjunction with the accompanying drawings, in

which:

FIG. 1 is a view useful for explaining a group control system for elevators; v

FIG. 2 is a view useful for explaining the principle of the present invention for discriminating traffic demand signals into several demand conditions;

FIGS. 3 and 4 are block diagrams showing the functions of an embodiment of the present invention, respectively;

FIG. 5 is a circuit diagram showing an example of distance function generator;

FIG. 6a is a block diagram showing an example of maximum selector;

FIG. 6b is an input to output characteristic diagram in a comparator;

FIG. 7 is a block diagram showing an example of minimum selector;

I detector;

FIG. 13 is a circuit diagram showing an example of the first pattern classifier; v

FIG. 14' is a view useful for explaining the prototype point thereof; 1

FIG. 15 is a circuit diagram showing an example of the second pattern classifier;

FIG. 16 is a view useful for explaining .the prototype point thereof;

FIG. 17 is a circuit diagram showing an example of the third pattern classifier;

FIG. 18 isaview useful for explaining the prototype point thereof; FIG. 19 is a view showing an example of the logic means;

FIG. 20 is a view showing another example of the pattern classifier; and

FIG. 21 is a view showing another example of the discriminant function generator.

The operational principle of the present invention is based on such an idea that a given demand signal is classified by using a discriminant function gi(X). Now, the traffic demand for juxtaposed elevators is grasped as an I-parameter demand signal X consisting of a plurality of factors of traffic demand X X X,, and an optimum discriminant function gi(X) is determined for each-demand condition. The term l-parameter demand signal X refers to a signal representing l-variable parameters. The term I-variable parameters" represents the function having any number of parameters. For example, where i=2, the l-variable parameters" represent a pair of variables corresponding to a two-dimensional space or plane. Furthermore, where i=3, a cubic or a three-dimensional volume having length, width and height may be used to represent the three parameters. For a set of parameters having 1 equal to or greater than 4, the function cannot be humanly visualized, but is a physical entity which can exist theoretically. Now assume that the discriminant function gi,,(X) which belongs to the demand signal i,, represents the largest value in all discriminant functions gi (X). In the above case, the traffic demand for'the elevators at that time is determined as the demand condition i,,.

The discriminant function representing a maximum value described above refers to one which represents a substantially maximum value. Thus, even if superficially an attempt is being made to select a discriminant function which represents a minimum value, substantially a discriminant function which represents a maximum value is being selected. In case g;, (X)

of functions g,(-X), g (X) and 3 (X) represents a maximum value then the function representing a minimum value will be -'g,,( X on the assumption that the signs of these functions are negative. Conversely, if g,(X) of the functions g,(X), 82(X )and g (X) represents a minimum value, then g,(X) will represent a maximum value, on the assumption that the signs of these functions are negative.

Various types of functions may be employed as discriminant functions, but inthe following discussion, description will be made of ones of which the prototype points are set up for the respective demand conditions for the sake of simplicity. The term fprototype point" corresponds to a representative point, which is employed to divide an n-parameter function (where n is an arbitrary number) into more than two categories. That is, in the embodiments described herein, traffic demands for the elevators are continuously detected by means of various detectors, detection is made of that one of the prototype points to which the demand signal X is closest which consists of a plurality of factors of traffic demand resulting from said continuous detection, and the demand condition to which the prototype point thus detected is dealt with as the demand condition at that time. Thus, it is possible to set up the prototype points so as to cope with the status of each individual building and yet freely adjust them for all the factors of traffic demand.

Assume now that the demand signal is represented by an parameter vector X=(X,, X X X,) consisting of I detection elements, and that the demand conditions are given by 0P 0P OP I The detection elements described above may include the following. The demand conditions are as already described. Factors of traffic demand: t

A. Weight of the elevator cage during the up" trip I B. Weight of the elevator cage during the down" trip C. Number of up hall calls D. Number of down hall calls E. Others By dividing the l-variable parameter function in which the elevator demand signal X represented by the'plural factors of trafiic demand X X X, X, into R regions in accordance with the demand conditions (such division is effected so that In this case, it may be conceivable to continuously and precisely section the l-variable parameters in which the demand signal X is present into R demand conditions, but this is not practical.

Here, description will be made of the case where an attempt is made to operate the elevator group in accordance with an operating program which corresponds to the demand condition to which belongs that one of a suitable number of prototype points set up for each demand condition which is closest to the demand signal X, with said demand condition as the op-' timum demand condition at that time.

Assume that the demand signal X is two-dimensional or has two variable parameters, that the demand conditions are represented by OP, to OR, respectively, that the prototype points of the demand condition OP are represented by I", and Pi", that the prototype points of the demand condition P0 are represented by P and P and that the prototype points of the demand condition 0P are represented by P and P as shown in FIG. 2. Then, in the case where the demand signal X is as shown in the drawing, it is closest to the prototype point I; so that the optimum demand condition is determined to correspond to 0P and thus the operation is performed in accordance with the operating program. In the cases where the demand signal represents three-parameters or higher, too, what has been described just above holds true. Generally, this may be explained as follows:

Assume that the following S prototype points are set up in an l-parameter vector set constituting the demand signal X.

I P P Pf'flrototype points belonging to the first-demand condition) P P5", Pi, Pk flrototype points belonging to the second demand condition) PE; P PA, Pk ilrototype points belonging to the Rth demand condition) S=L,+L +L +L,,

In order to determine the optimum demand condition,

therefore, it is necessary to select that one of the S prototype points P: which is closest to the distance between the demand signal X and each of the S prototype points.

The square of the distance between the demand signal X 5 and the prototype point 1 is represented by =X.X P j., l,1 j.p .i 1)10 where a" indicates inner produ ct. From this, it will be seen that the prototype point closest to the demand signal X is the one where the 2( i 'Xi i r Then, the term g,(X) which represents a maximum value is sought from L, functions.

j=1 2, 3, L, i=1, 2, 3, R By selecting the following term g (X) lo( X)=MaX 8:00 which represents a maximum value from the R distance functions which are selected as one for each demand condition, the i th demand condition to which this term belongs becomes the optimum demand condition in this case. Assume that the hth oneP (i=f, L,=h) of the prototype points belonging to the jth demand condition has the short space distance with respect to the demand signal X. Then, the hth one of the prototype points belonging to the jth demand condition is closest to X, so that the distance function of that prototype point or 4( )=P 'XaP;-Py- (5) g V V H 1 becomes maximum. Thus, g','(X) is selected. in the case of other demand conditions, too, the prototype points each of which represents a maximum value are similarly selected.

In this way R prototype points of the respective demand conditions (each prototype point is closest to X among the prototype points belonging to each demand condition.) are selected.

it will be apparent that g,(X) belonging to the jth demand condition becomes maximum since that one of the functions 81(X) which represents the greatest value is closest to X.

Thus, it is judgedthat the prototype point which is closest to the demand signal X is P,, and that the demand condition is the fth one. Consequently, the mode of operation of the elevator group is switched to the fth operating program.

In the foregoing, description has been made of the case where use is made of the following function in order to effect discrimination with respect to the prototype point closest to X g;'(X)=P{-X L-P ,I-P I However, it is also possible to use any other ftinctionl As described above, it is essential that at least one prototype point which represents each demand condition be set up in an l-paramctcr vector set in which the demand signal X is present, and that the prototype one which has the minimum function distance with respect to X'be selected. it has also been described that this can be achieved by effecting operations represented by equations (2), (3) and (4).

Next, description will he made of the case where each demmul tolttllllott lm'lmlnnonc prototype point (this case corresponds toj--l for the sake of simplicity.

Generally, a vector inner product is represented by P X,+P X +P,X ..+P,x,

on the assumption that the components of the vector P are P,

P P P, and those of the vector X are X X X X,.

Thus, it is assumed that the components of the demand signalX are X X X X, and that the components of the prototype point P, of the ith demand condition are P P P P11, then the following equation holds true P 'X =PHX1+PHXZ+ +MEHX I... Similarly, the following equation holds true Therefore, equation (2) is reduced to Since are all correspond to the components of the prototype point P each of them assumes a predetermined constant value. By representing these terms by W which is called weight, the distance function of the prototype point P, and X is sought from the following equation:

g,(X)=W W +W, X +W Xi+W +l (Hitwhcrc and k=l, 2, 3, I. Generally, if it is assumed that there are demand conditions, then the distance functions of R prototype points and vector X are respectively given by Thus, by seeking that one of the R distance functions which represents the maximum value, it is possible to determine the demand condition to be sought to which the thus sought distance function belongs.

For example, on the assumption that the demand signal is represented by X, if the following equation v becomes maximum, then the 5th demand conditiori (i=5) is selected so that the mode of operation of the elevators is switched to the fifth operating program corresponding to the selected demand condition. I 7

Consider the case where the number ofprototype points belonging to each demand condition is j which is 2 or more. For example, if it is assumed that the jth prototype point in the ith demand condition is represented by P and that the components of this prototype point are respectively represented by P5, PR P{,, then the distance function giUi) of the prototype point and vector X is given by the following equation as in the foregoing case:

where IVMJ= M It I, 2,3,...1

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US3073417 *Dec 23, 1959Jan 15, 1963Otis Elevator CoElevator dispatching and control system
US3080944 *Jul 6, 1960Mar 12, 1963Toledo Scale CorpElevator controls
US3292736 *May 16, 1961Dec 20, 1966Westinghouse Electric CorpElevator system with sequence for selecting an available car and expedited service for main floor
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US4084661 *May 7, 1974Apr 18, 1978Westinghouse Electric CorporationElevator system
US4677577 *Apr 21, 1986Jun 30, 1987Mitsubishi Denki Kabushiki KaishaApparatus for statistically processing elevator traffic information
US4760896 *Sep 29, 1987Aug 2, 1988Kabushiki Kaisha ToshibaApparatus for performing group control on elevators
EP0968953A1 *Jan 19, 1998Jan 5, 2000Mitsubishi Denki Kabushiki KaishaManagement controller of elevators
WO2013036225A1 *Sep 8, 2011Mar 14, 2013Otis Elevator CompanyElevator system with dynamic traffic profile solutions
Classifications
U.S. Classification187/382, 187/391
International ClassificationB66B1/18, B66B1/20
Cooperative ClassificationB66B1/20
European ClassificationB66B1/20