JPH0816842B2 - Ambiguous inference control command decision method - Google Patents

Description

The present invention relates to a control command determination method, and more particularly to a fuzzy inference control command determination method suitable for fixed position stop control of a vehicle driven by a discrete control input such as a notch.

[Conventional technology]

FIG. 1 is a block diagram showing an example of a conventional automatic train control device for automatically stopping a train at a target position, and FIG. 2 is an operation explanatory diagram showing its speed pattern. In FIG. 1, 1 is a speed generator, 2 is a speed calculation circuit, 3 is a distance integration circuit, 4 is a ground element, 5 is a point detector, 6 is a pattern generator, 7 is a comparator, and 8 is a brake controller. And 9 has shown the brake device.

In the above-mentioned conventional fixed position stop control, the vehicle uses a ground detector 4 installed at a point P (point P) located a predetermined distance l before the stop target point O (point O) as a reference, and a point detector on the vehicle. 5 to detect. The distance integration circuit 3 counts the pulses from the speed generator 1 according to the point detector detection signal from the point detector 5, and outputs the distance S traveled by the vehicle from the point P. In addition, the pattern generator 6 is the point detector 5
From the point detection signal from the
From its own memory and using a predetermined deceleration β _P The target speed pattern V _{P according} to the following equation is output.

The speed calculation circuit 2 calculates the actual speed V _{T of the} vehicle from the pulse of the speed generator 1. The actual speed V _T and the pattern generator 6
The comparator 7 compares the target speed pattern V _P output from the. The brake device 9 is proportionally controlled via the brake control device 8 based on the output signal (V _T −V _P ) of the comparator 7 so that the actual speed V _T of the vehicle follows the target speed pattern V _P. Controls to stop near the stop target point O.

However, in the above-mentioned method, the stop accuracy may be deteriorated due to the occurrence of a steady deviation following the target speed pattern V _P due to the fluctuation of the braking force of the brake device 9 and the fluctuation of the external force due to the gradient of the track. Further, in order to prevent the occurrence of the following steady-state deviation, if integral control is performed in addition to proportional control of the brake device 9, a control command (notch)
Another problem occurs that the number of changes increases and the riding comfort of the vehicle deteriorates.

[Object of the Invention]

The present invention has been made in view of the above circumstances, and an object of the present invention is to reduce the number of times of changing a control command (notch) as much as possible even when the vehicle is stopped at the target point as described above. Another object of the present invention is to provide a control command determination method using fuzzy inference that can realize good driving performance.

[General Description of Invention]

In order to achieve the above object, the present invention adopts the following configurations.

A control command determination method for determining a control command according to a state of a control target using a microprocessor, and whether a control purpose can be achieved as a result of inputting a plurality of control commands for the control target and each of the control commands in advance The decision items shown below are associated with each other, and a control rule is defined in which each control command and decision item combination is used as a condition part, and a control command to be executed corresponding to the condition part is used as a conclusion part. Prepare a membership function that represents the degree of ambiguity that indicates how well the control result matches the control purpose.
Prior to determining the control command, the control result when each control command is given to a plurality of different control commands is predicted from the motion model of the control target and the given control command,
Each predicted control result is converted into an ambiguous amount according to the above-mentioned membership function, one of a plurality of control commands is selected by comparing the magnitude of the converted ambiguous amount, and the control target is selected by the selected control command. Is controlled. Further, the following processing may be performed by the microprocessor.

In a control command method for determining a control command according to the state of a control target using a microprocessor, a determination indicating whether or not a control purpose can be achieved as a result of inputting each of a plurality of control commands for the control target and the control command in advance In addition to associating items with each other, define a control law that uses the combination of each control command and the determination item as the condition part, and the control command to be executed corresponding to the condition part as the conclusion part, and respond the control result to the determination item. By providing a membership function that represents the degree to which the control result matches the control purpose as an ambiguous amount, select the control command for the control target, and perform the control when the control is performed by the selected control command. The error with the purpose is obtained, and based on the error and the state of the controlled object, the control start margin time which is the time until the control by the control command is started is obtained. If the control start time is less than the specified time, the control result when the control commands are given for a plurality of different control commands prior to the determination of the control command is the motion model of the control target and the given control command. From each of the plurality of control commands by converting the predicted control results into fuzzy quantities according to the membership function and comparing the magnitude of the converted fuzzy quantities.
One is selected and the control target is controlled by the selected control command. In order to reduce the number of control command changes and perform smooth control, the above control result is predicted for the current control command and the control command that has changed it by a predetermined amount, and the current control command is the most satisfactory. The same control command may be maintained as it is as long as a possible control result is obtained.

Example of Invention

FIG. 3 is a block diagram of a vehicle fixed-position stop control system showing an embodiment of the present invention, and the same components as those shown in FIG. 1 are designated by the same reference numerals. Also, 2'is a speed generator 1
The speed of the vehicle (actual speed) V _T by the pulse for the past 1 second of
Velocity acceleration computing device for computing acceleration α _T , 3 ′ is a remaining distance computing device for computing L ₁ which is the distance between the current position (x ₁ ) of the vehicle and the target stop point O, and 10 is the current vehicle speed Using V _T and the distance L ₁ to the stop target point O and the deceleration β _P that is predetermined corresponding to the preselected notch, L ₂ = L ₁ −V _T ² /(7.2·β _P ) ... (2) A predetermined notch stop error calculation device for calculating a stop error L ₂ when braking is performed by the notch previously selected by the formula (2), 11 is the stop error L ₂ and the current speed V _T
From this, the brake control start margin time ΔT when traveling at the current speed is calculated by the formula ΔT = 3.6L ₂ / V _T (3), and the margin time ΔT becomes a fixed time (for example, 1 second) or less. In this case, the brake control start command calculation device outputs a brake control start signal for stopping at a fixed position.

The 12Z, 12P, and 12N show the stop accuracy predicted when the vehicle is operated under different control commands, for example, when the notch selected in advance is held, when one notch is added, and when one notch is subtracted. A stop precision ambiguous amount calculation device for calculating a vague amount, a notch change suppressing device 13 for generating a signal M for suppressing a notch change only for a predetermined period after the notch change, for example, 1 second, and a stop accuracy ambiguous amount arithmetic device Vague inference is performed based on the output, and the notch command N should be output when notch change is not suppressed.
A control notch calculating device for calculating (t), and a brake command calculating device 15 for intermittently outputting the output of the control notch calculating device 14 according to the output of the brake control start command calculating device 11.

The operation of the apparatus of this embodiment will be described below with reference to FIGS. 4 and 5.

FIG. 4 is a diagram for explaining the operation of determining the brake control start time for the fixed position stop control, and the point P where the vehicle is placed a predetermined distance 1 from the target stop point O.
And shows the situation at x ₁ . The predetermined notch stop error calculation device 10 described above, determine the stop error L _2, this and from the current speed V _T, seeking brake start time margin ΔT by the brake control start command calculation unit 11, the brake start margin A brake control start command is output when the time ΔT becomes a certain time (for example, 1 second) or less.

FIG. 5 is a diagram for explaining the operation of selecting a brake notch for stopping at a fixed position. In the stop precision ambiguous amount calculation devices 12Z, 12P, 12N, the distance L ₁ to the stop target point obtained by the remaining distance calculation device 3'indicating the state quantity of the vehicle and the current obtained by the velocity acceleration calculation device 2 ' Speed V
_T1 , acceleration α _T1 and predetermined acceleration per notch Δα
From (for example, 0.5 Km / h / sec), the scalar amounts L _Z , L _P , and L _N of the stop accuracy, which is the control purpose, are calculated by the following equation (4).

Here, it is assumed that the ambiguous amount of stop accuracy is composed of two ambiguous base sets: ( ₁ ) can be stopped within a tolerance (μ ₁ ) and (2) can be accurately stopped at a stop target position (μ ₂ ). A membership function that represents these fuzzy sets is defined as follows, for example. x is the stop error (m) and the allowable error is 0.
If the distance is 5 m, the membership function μ ₁ (x) of the fuzzy platform set that can be stopped within the above tolerance is And the membership function μ ₂ (x) of the fuzzy base set that can accurately stop at the stop target position is Is defined as

The output of each of the stopping accuracy ambiguous amount computing devices 12Z, 12P, 12N is an ambiguous amount consisting of a pair of values of the above two membership functions μ ₁ (x) and μ ₂ (x) with respect to the predicted stopping accuracy. It represents. That is, from the scalar amount L _Z , L _P , L _{N of the} stop accuracy, if the current notch is held,
Also, the ambiguous amounts _Z , _P , _N of the stop accuracy when the notch is changed by ± 1 notch from the current value are obtained as values represented as follows.

Based on the outputs _Z , _P , _N from the stop precision fuzzy amount computing devices 12Z, 12P, 12N, the control notch computing device 14 may control control rules (1) to ((1) to ( The control command to be output, that is, the notch is selected based on the fuzzy reasoning according to 3).

(1) If the current notch can stop within the allowable range,
Keep the current notch.

(2) If it is possible to accurately stop at the stop target position with +1 notch, increase by 1 notch.

(3) If it is possible to accurately stop at the stop target position with -1 notch, decrease by 1 notch.

The specific notch selection based on the fuzzy reasoning is performed by selecting the value of the membership function corresponding to the determination items defined in the respective conditional parts of the control rules (1) to (3) (in the embodiment, respectively, C _This is done by selecting the one that maximizes _Z1 , C _P2 , C _N2 ). Note that this control is performed when there is no notch change suppression signal M from the notch change suppression device 13.

In the present embodiment, the first consideration is that the target stop point can be stopped within the allowable error, and the notch is changed only when the stop can be stopped with higher accuracy by changing the ± 1 notch. It is possible to realize a fixed position stop control that is safe and has little change in notch and is comfortable to ride.

In the above embodiment, an example in which the various arithmetic units 10 to 15 are configured by hardware has been shown, but a microcomputer may be used for the arithmetic. An example of the program processing flow in this case is shown in FIG. This program has a function of obtaining the notch command N (t) based on the vehicle speed V _T , acceleration α _T , and the distance L ₁ to the stop target point.

Further, in the above-mentioned embodiment, the membership function of the fuzzy base set for defining the fuzzy amount of the stopping system, which is the control purpose, is defined as in the formulas (5) and (6). It is not limited to. For example, as shown in FIG. 7, the vehicle speed is a predetermined speed V _C (for example,
10 Km / h) or more, the point Q, which is 1 '(for example, 2 m) before the final stop target point O, is targeted, and the point O is targeted when the speed becomes V _C or less. It may be changed to.

In the fuzzy reasoning in the above embodiment,
Although the number of notches is increased or decreased by ± 1 notch, it goes without saying that more notch changes may be added to the selection range by increasing the number of control rules.

〔The invention's effect〕

As described above, according to the present invention, in each condition part of a plurality of control rules consisting of a condition part indicating a matter to be judged and a conclusion part indicating a control command to be executed, the matter to be judged and its premise The relationship between the control command is defined, and before the control command is determined, the control result of the control target when each control command is given for a plurality of different control commands is predicted, and the predicted control result is calculated. A control for converting a fuzzy amount according to the definition of the conditional part of the control law, and executing a control command indicated by the conclusion part of one of the plurality of control laws selected based on the fuzzy amount. It outputs it as a command. Therefore, if the evaluation of the predicted control result with respect to the current control command is better than the others, the current control command is continued as it is, the change of the control command is reduced, and in particular, the device of the present invention is used to automatically drive the vehicle. When applied to, it is possible to control the stop comfortably in the vehicle.

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of a conventional fixed position vehicle stop control device, FIG. 2 is an operation explanatory diagram showing its speed pattern, FIG. 3 is a block diagram showing one embodiment of the present invention, FIG. FIG. 5 is a diagram for explaining the operation, FIG. 6 is a processing flowchart showing another embodiment of the present invention, and FIG. 7 is a diagram showing a speed pattern in still another embodiment of the present invention. . 1: Speed generator, 2 ': Speed acceleration calculation device, 3': Remaining distance calculation device, 4: Ground element, 5: Point detector, 9: Brake device,
10: Stop error calculation device, 11: Brake start command device, 12Z,
12P, 12N: Ambiguous amount calculation device, 13: Notch change suppression device, 14: Control notch calculation device, 15: Brake command calculation device.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-58-190204 (JP, A) Terano, “Introduction to Applied Fuzzy Systems” September 20, 1989, Ohmsha, PP. 110-122

Claims (1)

[Claims] 1. A control command method for determining a control command according to a state of a controlled object by using a microprocessor, wherein a plurality of control commands for the controlled object and each of the control commands are input in advance, and a control purpose is determined. In addition to associating the judgment items that indicate whether they can be achieved, with the combination of each of the above-mentioned control commands and the above-mentioned judgment items as the condition part, define the control rule that the control command to be executed corresponding to the above condition part is the conclusion part. Then, prepare a membership function that represents the above control result as an ambiguous amount that indicates how well the control result matches the control purpose by selecting the control command for the control target. The control command is calculated based on the error and the state of the controlled object when the control target is controlled by the selected control command. When the control start margin time, which is the time until the control is started, is obtained and the control start margin time becomes less than or equal to the predetermined time, each control command is output for a plurality of different control commands prior to the determination of the control command. Is predicted from the motion model of the controlled object and the given control command, each predicted control result is converted into an ambiguous amount according to the membership relationship, and the converted ambiguous amount is small or large. The fuzzy inference control command determination method is characterized in that one of the plurality of control commands is selected by comparing and the control target is controlled by the selected control command.