US 3844242 A
Apparatus is provided for dynamically maintaining a ship in given position by the automatic control of propellers which respond to signals generated by position pick-offs and accelerometers. The accelerometers take into account longitudinal, lateral and vertical accelerations, when necessary, and an accelerometer is provided for angular acceleration in heading as well as, if necessary, rolling and pitching. Signals from the accelerometers are sent to servo systems which generate propeller thrusts which counteract the disturbances.
Description (OCR text may contain errors)
United States Patent [1 Sernatinger et al.
1 APPARATUS FOR AUTOMATIC DYNAMIC POSITIONING AND STEERING SYSTEMS Inventors: Franz Sernatinger, Mornac;
Maurice Abad, LIsle DEspagnac, LQL I France.
Assignee: Etat Francais, Paris, France Filed: Sept. 20, 1972 Appl. No.: 290,516
 Foreign Application Priority Data Sept. 21, 1971 France 71.33836 US. Cl. 114/144 B, 318/588 Int. Cl B63h 25/42 Field of Search 114/16 E, 144 B, 121, 122, 114/16 R; 244/77 G, 77 D, 77 R, 17.13, 50; 318/588, 489; 235/1502; l80/79.l, 79.2
References Cited UNITED STATES PATENTS 3/1967 Berne 114/144 B Oct. 29, 1974 3,318,275 5/1967 Field 114/122 X 3,481,299 12/1969 Horn... ll4/l44 B 3,547,381 12/1970 Shaw 244/77 D Primary Examiner-Trygve M. Blix Assistant Examiner-Stephen G. Kunin Attorney, Agent, or FirmWaters, Roditi, Schwartz & Nissen 5 7 ABSTRACT 7 Claims, 12 Drawing Figures 1 came.
NE TWD/e ACCELLEEA T/OIV pas/r/a/v 3 Z 04 250: mam N'fWd can v. J" J PK -0 2 20 /9 X Ma a/25w:- yo POs/T/O/V APPARATUS FOR AUTOMATIC DYNAMIC POSITIONING AND STEERING SYSTEMS FIELD OF INVENTION The present invention relates to automatic dynamic positioning and steering systems intended for surface ships and submarine craft and the like.
BACKGROUND It is known that dynamic positioning consists of maintaining a ship in a given position and heading and, also at a determined immersion when a submarine craft is concerned, by means of propellers only, the orientation and/or thrust intensity of which is adjusted to oppose external forces which tend to move the craft away from the desired position.
The propulsion systems in use include at least two swivelling thrust direction propellers such as outboard motors or else two cycloidal type propellers with vertical blades and axes of rotation with one of such propellers being located forward and the other aft on the associated ship. Another means which is used consists of housing propellers delivering adjustable and reversing thrust in tunnels running through the ship in the vicinity of its bow and stem, respectively, permitting application of a lateral thrust and a rotational moment as opposed to the thrust between the front and the rear. The longitudinal thrust is provided by one or more propellers located at the rear, these propellers also permitting displacement of the ship between positioning stations.
The position in the horizontal plane is sensed with respect to a fixed point defined on the bottom of the sea and in a system of coordinates fixed or related to the moving body, by means of a position pick-off such as, for instance, a ratio electric system, with beacons or transmitters located ashore, an acoustic system such as a sonar or ultrasonic transponder beacon, or a cable stretched between the ship and a mooring placed on the bottom, the slanting angle of which is measured with respect to the vertical.
The vehicle direction is measured with a magnetic compass or a gyro compass. The depth in the case of a submarine craft is determined with respect to the surface by measuring the pressure or with respect to the bottom by an ultrasonic depth-finder. The values measured by these pick-offs are compared with reference values set and defined in a system of reference axes such as for instance (x,,, y,,) for the position in the horizontal plane, 0., for heading, and Z for the immersion.
Any change in these reference values signifies the associated ship's displacement and represents a steering function.
The propeller control from the comparision of the reference values with those measured by the position pick-offs described above forms a servo-system which performs the automatic dynamic positioning such as it has been defined.
The existing servo-systems feature a number of drawbacks which limit their performance characteristics notably as concerns positioning accuracy and operating flexibility. The major drawback is related to the type of position pick-offs, especially those based on the use of ultrasonics. As a matter of fact, these pick-offs feature quite a long response time resulting from the travelling speed of sound in water and they may be subject to momentary losses of the incoming signal. In addition, the reception is affected by noises coming, for example, from the propellers and ships movements.
The compromise which is then sought for in order to determine the corrector networks (such as integral and differential corrector networks) according to the conventional theory of servo-systems entails undesired command signals on the propellers, worsening the mechanical wear and fuel consumption while featuring poor performance characteristics under transient conditions.
In addition, important changes in the ships response to applied forces may occur for various reasons: variations in weight and center of gravity location in proportion to the loading; bouyancy for submarine craft varying with temperature and water salt contents, immersion duration as a consequence of the gradual disappearance of the air bubbles caught in the upper works, immersion depth because of the volume varying as the pressure; and propeller thrust hydrodynamically deviated by current. Such changes in ships response make it necessary to further reduce the performance characteristics or to leave at the operating personnels disposal adjustment controls for the said networks in order to maintain normal servo-system stability margins, thereby increasing duties and the risks of errors.
Also, this poor response to transient conditions makes steering very difficult, i.e., the change in reference point defined above.
SUMMARY OF THE INVENTION A first object of the invention is to improve the posi tioning accuracy notably under transient conditions associated with disturbances.
A second object of the invention is to obtain the first result whatever the type of position pick-off used and in spite of important ship or craft weight variations.
Another object of the invention is to facilitate steermg.
The automatic dynamic positioning and steering system according to the invention incorporates accelerometers in addition to one of the position pick-offs already mentioned. For example, three linear accelerometers supply signals proportional to longitudinal, lateral and vertical accelerations, respectively. The last measurement is obviously not essential when a surface ship is concerned.
An angular accelerometer generates a signal proportional to the angular acceleration in heading.
' Generally, a ship is hydrodynamically stable in rolling and in pitching and does not incorporate any servosystem to maneuver relative to these axes. This does not limit the scope of the invention which can be extended, in particular, to the case of a submarine craft through the use of accelerometers measuring angular, rolling and pitching accelerations. The signals issuing from such accelerometers are reinjected to the associated servo-system inputs in order to control the propellers and thus achieve an accelerometer feedback. This feedback imposes a propeller thrust which tends to null out any acceleration, i.e., any vehicle movement, thereby achieving dynamic positioning.
However, it is known that such pick-offs cannot be absolutely perfect in practice, as notably the measurement of very small accelerations is affected by an unavoidable shift of the unit zero, which shift is also called drift. This drift would result in a slow displacement of the mean vehicle position. The position pick-offs previously described permit such a deviation to be detected and to be nulled out by also operating the propellers.
Combining both types of pick-offs provides simultaneously, on the one hand, an accelerometer channel with a wide passband capable of efficiently opposing fast disturbances or of following without any delay changes in reference points and, on the other hand, a position channel'with a passband as narrow as necessary to efficiently filter out the position pick-off signals. This channel provides mean position accuracy and slow-variation disturbance compensation. Also, the passbandsoverlap so that the accelerometer channel provides for positioning, i.e., a position memory function, during the momentary absences of a signal in the position pick-off.
A channel is known as an anticipation channel, thus called because it applies control signals to the propellers which are generated before the pick-offs previously described have been able to evidence a displacement of the vehicle. As a matter of fact, the thrust and moments applied to a vehicle according to direction and speed of the wind, current and squall, respectively, are known with quite a good approximation, either from computation or from measurement made on a model in a model basin or wind tunnel. Pick-offs include such as anemometer, current meter, swell meter, and supply signals proportional to the speed and direction of the disturbing causes. Such data are processed according to an analog or digital computation method in order to obtain electrical voltages proportional to the disturbances computed from known relations, which voltages are applied to the propeller control to obtain the appropriate thrusts in opposite directions.
With the disturbances thus compensated for by this anticipation channel, the acceleration and position loops will have only to oppose the differences between the actual disturbances and the computed disturbances since the formulae deducted from model basin testing for instance are not always actually rigorous. However, dynamic and static positioning errors will be substantially reduced.
This system has been improved by assigning an amplitude schedule to the propeller control signals which is the inverse of the propeller thrust response so that the resulting thrust becomes proportional to the com-' mand signal applied.
The system according to the invention also incorporates an automatic gain and phase adjustment for the corrector networks arranged in the acceleration and position channels so as to satisfy the stability and accuracy criteria in compliance with the conventional theory of the servo-systems. For this purpose, periodic signals are applied to the propeller control with quite a low amplitude in order to avoid interfering with equipment operation. The accelerometers can, however, sense the resulting vehicle movement and the correlation between these measured signals and applied signals gives an indication of the vehicle inertia response relative to the thrust.
BRIEF DESCRIPTION OF THE DRAWING Other characteristics and advantages of the invention will be evidenced by illustrating it with a non-limitative but illustrative embodiment and by reference to the appended drawings wherein:
FIG. 1 represents a propeller tunnel arrangement, this arrangement proving to be the most convenient as an example of force breakdown;
FIG. 2 is a block diagram of the geometrical generation of propeller thrust command signals in the horizontal plane from the desired forces and moments along the major moving body axes;
FIG. 3 is a block diagram of the longitudinal channel servo-system, the lateral channel and moment channel servo-systems being similar;
FIG. 4 is a chart which illustrates propeller thrust in proportion to command signal;
FIG. 5 is a chart which indicates the amplitude response of the thrust control linearization circuit;
FIG. 6 is a diagrammatic perspective view of a sam ple accelerometer arrangement on a horizontal platform;
FIG. 7 is a block diagram of the compensation means for the angular movement effects on linear accelerations;
FIG. 8 is a graphic representation which summarizes the acceleration distribution in the transverse plane;
FIG. 9 is a block diagram of a self-matching channel;
FIG. 11 is a block diagram illustration of a sample computation of the anticipation terms; and
FIG. 12 is a block diagram of heading slaving to the disturbance direction.
DETAILED DESCRIPTION In FIG. I, aft propeller 4 delivers a longitudinal thrust fx, positive or negative along axis xx. Thrust f,,, of forward propeller 2 is parallel to axis yy' and is oriented to the right or to the left as is thrust f of aft transverse propeller 3. This arrangement in the horizontal plane is obviously applicable to the vertical plane for a submarine craft. 7
The conventional breakdown of forces permits the expression of propeller thrusts versus forces defined in the rectangular axis system (xx', yy) to be expressd as F and F,,, and moment C about axis 22' perpendicular to the preceding ones and passing through the ships center of gravity 0.
l, and 1 are respectively the distances from the forward and aft lateral propellers to the center of gravity 0.
These relations are applied according to FIG. 2 with a view to obtaining propeller command signals f,, f and f,,,, respectively, from thrusts F, and F and desired moment C. These signals are presented in the form of electrical signals with a suitable scale factor. Amplifier circuit 5 inserts a gain proportional to 1 just as amplifier 6 multiplies F u by l,. Adder 7 makes the sum of the terms F l obtained previously, and C, by multiplying the result by a coefficient analogous to 1/1, l whereas subtractor 8 makes the difference F I, C with a gain proportional to the same coefficient l/l, +1
This breakdown permits afterwards to reason by individually examining the longitudinal, lateral and moment channels, respectively. Each of the channels then features the same structure such as that shown in the block diagram of FIG. 3 which corresponds to the longitudinal channel as an example.
Block 9, FIG. 3, represents the vehicle with its transfer function, the output being its position, the input being the resultant on the one hand of disturbing forces F due to the wind, current and swell and, on the other hand, thrust F produced along axis xx by propellers l0.
Block 13 which does not correspond to a material realization, symbolizes in the schematic diagram the physical system of disturbing forces and moments.
As in most applications of this type ofpropeller which is available on the market and on which no further details will be given, command signal S acts on the screw propeller pitch. The thrust then is not proportional to the command and has in fact the appearance of curve 27, FIG. 4. This non-linearity may be detrimental to servo-system stability. Amplifier 11, FIG. 3 makes the thrust linear by amplifying command signal S with an amplitude response such as that shown by curve 30, FIG. 5, which is the inverse of curve 27. The
thrust of the corresponding propeller is thus proportional to command signal S Zero-slope segments 31 of curve 30 determine a limitation of the thrust at a selected value lower than the maximum permissible power, as shown by curve 28.
According to an advantageous mode of realization, this linearization is achieved by an analog function generator. According to another mode of realization, when a digital computer is employed, curve 30-31 can be approximated by a table of stored discrete values.
Amplifier II incorporates, in addition to the said function generator, a linear unit the gain of which can be adjusted as described afterwards, in order to adjust the slope of a linearized curve between 28 and 29. As the propeller thrust is applied to the ships inertia, it results that the ships inertia acceleration, for slow movements and below the limitation threshold is proportional to the direct channel input signal 31 which direct channel incorporates items 9, l0 and I1 defined above.
Command signal S, itself results from the combination, according to the invention, of the signals generated by the various channels identified in FIG. 3 by the arrow lines. Adder 12 makes a sum of these signals.
Accelerometer assembly 14 (see, e.g., Locks Guidance published by Van Nostrand chapters 9-5 and 9-6 Pgs. 337 ff) measures the longitudinal vehicle acceleration. It notably incorporates a linear accelerometer 33 mounted on a platform 32 kept horizontal by known means. An advantageous mode of realization shown in FIG. 6, consists of using the inner gimbal ring of a vertical gyro 36. On this same horizontal platform 32 are then also mounted linear accelerometer 34 oriented along axis yy' and angular accelerometer 35 the sensitive axis of which is parallel to axis 22'.
Generally, this platform cannot be placed in the center of rotation which is also the quiet point" of the ship. As a result, during rolling and pitching movements, a tangential acceleration occurs on the platform. The tangential acceleration projections onto axes xx and yy are detected by accelerometers 33 and 34, respectively. However, by installing the platform substantially straight above the quiet point at a distance d (distance between the accelerometer platform deck and ships center of rotation), the equations expressing the disturbances are greatly simplified.
According to the invention, a mode of compensation for these disturbances, the schematic diagram of which is shown in FIG. 7, consists in arranging two angular accelerometers 39 and 40 measuring the angular accelerations in rolling 0.x and pitching fly, respectively.
Integrators 41, 42 and 43 supply from the angular accelerations, angular rates 0,, Q and Q Multipliers 44, 45 and 46 square 0, and Q and the product 9,, (2 respectively. Adders 47, 48 and 49, weighed in proportion to said distance a, supply signals which represent the following terms:
These terms actually correspond to the projections of the spurious linear accelerations resulting from the angular movements onto the axes of a trihedron related to the ship with the platform installed in the manner described above.
These components are projected into the trihedron related to horizontal platform 32 by means of resolvers 37 and 38 which generate signals representative of the spurious accelerations:
v. 7, cos T+ (w sin R cos R) sin T 'Yu 'Yu CO5 R ')z Sin R which are subtracted from the values measured by accelerometers 33 and 34. The outputs from adders 50 and 51 thus correspond to the corrected longitudinal and transverse accelerations which are used in the acceleration and self-matching channels.
As a matter of fact, it can be noted that, in general, the product of the angular rates is small as compared with the angular acceleration. In addition, the rolling rate is in quadrature with the rolling angle, likewise the pitching rate is in quadrature with the pitching angle and this pitching angle does not exceed but a few degrees. According to the simplified formulae, the spurious accelerations are written as follows:
'y",,# Q, d cos R as illustrated by FIG. 8 for rolling.
According to FIG. 9, the signals supplied by angular accelerometers 39 and 40 are subtracted by adders 50 and 51 from those from linear accelerometers 33 and 34 with a weighing factor which corresponds to the multiplication by distance d, Qx additionally crossing resolver 37 to be multiplied by cos R.
Again referring to FIG. 3, the corrected accelerometer signal is applied to corrector network 15 of the proportional-integral type, then the loop is closed through adder 12. The parameters of network 15 are determined according to the conventional methods of computation for the servo-systems and obviously depend on the ships transfer function. The acceleration channel gain in open loop being G, the known formula giving the acceleration taken by the ships weight under the influence of a disturbing force F p is:
The acceleration loop has thus the effect of artificially multiplying the ships inertia by (l G) for small movements and in the band width of the accleration loop, proportionally reducing ships inertia and sensitivity to disturbances. In addition, the thrust linearization action by item 11 is completed by the servo-system in order to make the acceleration taken by the vehicle proportional to a command signal applied to the other inputs of adder 12.
Position pick-off 17 provides information on the vehicle deviation within the reference system to which this pick-off is related. The measured deviation is compared by adder 18 to a deviation x, set by means of adjustment resistor 19 which constitutes the reference position. Known coordinate converter 20 is used if necessary to obtain the position deviation components in the trihedron related to the ship when the position pick-off reference system in use is different. Corrector network 21 filters out the position error. The possibility of adjusting accelerometer feedback corrector network provides very wide adjustment capability for network 21 which can then be optimized according to the pickoff in service. Notably, an advantageous mode of realization for network 21 consists of using a Kalmann filter as recurrent filter. The filtered position error is'applied to adder 12 to close the loop through doublethrow switch 22. When this double-throw switch is placed in the up position according to the diagram of FIG. 3, the position loop is opened and the output from potentiometer 23 is applied to the servo-system. This potentiometer is hand-driven through a manual control 124 and is automatically reset to zero which enables the adjustment of the propeller thrust and vehicle steering. The accelerometer feedback makes this steering particularly flexible.
To achieve continuous gain and phase matching of the corrector networks in order to compensate for the vehicle response variations, the signal generated by accelerometer measuring assembly 14 is also applied to circuit assembly 16 the details of which are shown in FIG. 10. Oscillator 52 generates an AC. signal, a portion of which is injected into adder l2 and the frequency of which is selected close to the acceleration loop cutoff frequency. The modulation resulting from the thrust which is of the order of a few percent of the maximum value is sufficient to cause a ship's movement detectable by the accelerometer without impeding the operation in any manner.
The measured acceleration is applied to multipliers 53 and 54 which are also fed with the signal from oscillator 52, directly to one and after a 90 phase-shift for the other. After filtering through low-pass filters 55 and 56, in-phase component u and quadrature component v of the ships movement are obtained, respectively.
This mode of detection performs the correlation operation and removes all the signals which are not synchronous with the applied oscillation, notably those due to the swell. A known circuit incorporates resolver 57 (see, e.g., Chpts. 6-11, Pgs. 329ff Korn and Korn Electronic Analog Computers by McGraw-Hill) motor 59 and its servo-amplifier 58 and supplies moduls A V 14 v from components u and v and phase 4) Arc tg (u/v) of the synchronous acceleration. These values are compared with an amplitude and a phase preset by adjustment resistors 60 and 61, respectively, with the deviations amplified by 62 and 63 to control in the suitable direction the gain of the linear unit placed according to the preceding description in amplifier 11 as well as the time constants of corrector 15.
Again referring to FIG. 3, component 25 is a set of pick-offs of known types which measure the speed and direction of the wind, current and swell. The pick-off output signals are applied to the circuit assembly designated by component 26 which simulates physical system 13 (the whole of the surfaces of the superstructures and hull subject to the impact of wind, tide and swell).
The forces and moments resulting from the influence of a wind having a speed V, and a direction 111,, can be computed according to the known formulas:
F P SI) 1: 11) C05 d l, m) V2 p v v uv 1 0 C l/2 p S L V C sin 2111,,
where p is the air density, S, is a reference surface of the ships emerged part and L is the ship's length, C C,,,, and C, are the aerodynamic coefficients which will have been measured on a model in a wind tunnel. Similar formulae express the effects of current on the immersed part, with the hydrodynamic coefficients determined from testing on a model in a model basin. The swell influence is more complex but the average thrust can be approximated by relations of same form deducted from model basin testing.
FIG. 11 shows a construction diagram of assembly 26. This analog simulation enables the anticipation term generation process to be well illustrated but it is obviously not limitative and a digital computation for instance of these terms remains in conformance with the invention.
Speed V is squared by multiplier 64 with a scale factor which makes the voltage obtained homogeneous with U2 p S, V,, Resolver 65 driven by an angle ill, multiplies this voltage by cos ill, on the one hand and sin 111,, on the other hand.
The term in cos 111,. is itself multiplied by sin ill, by means of resolver 66 which gives a resultant proportional to sin 2 41,. Multipliers 67, 68 and 69 make then the product of these three terms with appropriate coefficients C C and C,,,. Now, it is known that these coefficients are not strictly constant but vary as direction lll Here, they are generated from angle 41,, by analog function generators 70, 71 and 72, respectively, which reproduce the curves deducted from model testmg.
The same operations are performed from parameters concerning current and swell by sub-assembly 76 (see components 64-72). The voltages obtained are added by adders 73, 74 and 75 to form anticipation command signals F F W and C which are respectively applied to adder 12 in each of the channels. The propellers so controlled then oppose equivalent disturbances F F and C,, applied to the ship.
Computation of external disturbing forces therefore permits the determination in an advantageous manner of the resultant direction of these forces. Components F and F up are applied to resolver 77 of FIG. 12, with the resolveritself slaved by amplifier 78 and motor 79 according to the known circuit already shown in FIG. 10. The cos output from the resolver represents total disturbance module F the angular position being angle \II Now, it is known that the thrust exerted by the wind or current is minimum when the ship is headed into this wind or current. This condition is met by using as the moment channel error signal not the deviation from set heading 0,, but the output from resolver 83 particularly interesting application concerns off-shore drilling ships and oceanographic ships. It can also be applied to submarine wreckage searching craft or to submarine craft operating on well heads, as the advantages according to the invention appear as well for the maintenance of a stationary point as during delicate approach or contact maneuvers.
These steering qualities are also interesting for maneuvering and drawing alongside large ships and especially alongside giant tankers.
1. A dynamic positioning and steering system for a surface or submarine vehicle including a plurality of propellers, said system comprising position pick-off means, and accelerometer means on said vehicle to detect longitudinal, transverse and vertical accelerations as well as angular accelerations of the vehicle and which include wide passband accelerometer feedback means which produce voltages which insure quick positioning and oppose high frequency components of the disturbances, said feedback means performing as a position memory in case of a momentary absence of position data, said accelerometer means thus limiting the position pick-off means action to very low frequencies and continuous components thereby making it possible to optimize pick-off filtering, an external disturbance correction channel including means for wind, swell and current measurements, means for the computation of forces and moments resulting from wind, swell and current, and servo-systems coupled to the latter said means in order to produce equivalent and opposite thrusts.
2. A system according to claim 1 comprising an automatic vehicle response correction means including means for the injection of a predetermined A.C. signal into said servo-systems, means for the correlation at said A.C. signal and the accelerometer means voltages in order to measure the resulting disturbance, and corrector network phase and gain adjustment means for the maintenance of a constant overall response.
3. A system according to claim 2 wherein the acceleration means includes response linearization means in series with the voltages for each propeller in addition to the accelerometer feedback means the amplitude response of which is inverse to the thrust response so that the resultant is nearly linear.
4. A system according to claim 1 comprising linear acceleration measuring means including roll and pitch angular acceleration detection means, and means for the computation of spurious terms resulting from vehicle angular movements and subtraction of these terms in order to obtain corrected longitudinal and lateral accelerations.
5. A system according to claim 1 comprising-anticipation channel means including means for wind, current and swell measurements, and means for the computation of resultant disturbance direction and slaving of the ship in rotation so that the ship is headed into this resultant direction, said slaving replacing repetition of a reference heading when the computed total thrust exceeds a given preset value.
6. A system according to claim 1 comprising switching means for removing the position pickoff means and replacing the pick-off means with a manual control which because of the accelerometer feedback means transmits acceleration command signals and permits particularly flexible steering.
7. A system according to claim 6 wherein the switching means placing the manual control into service is divided according to each channel so that steering can be manual on one or more channels while the other channels are slaved to the associated position pick-off means.
* l l l