Method and device for stopping vehicle at predetermined position

• US 5,018,689 A
• Filed: 03/14/1989
• Issued: 05/28/1991
• Est. Priority Date: 04/27/1982
• Status: Expired due to Term

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Miyamoto, Shoji, Ihara, Hirokazu, Yasunobu, Seiji
• Primary Examiner(s): Butler, Douglas C.
• Application Number:
  US07/322,991
• Time in Patent Office: 805 Days
• Field of Search: 364/426.05, 364/426.04, 364/426.01, 364/426.02, 364/426.03, 105/61, 180/244, 180/282, 303/93, 303/96, 303/97, 303/20, 303/100, 303/104, 303/105, 303/106, 303/109, 246/182 R, 246/182 A, 246/182 B, 246/182 C, 246/184
• US Class Current: 246/182.00B
• CPC Class Codes: B60L 15/00 Methods, circuits, or devices for controlling the traction-motor speed of electrically-propelled vehicles
• CPC Class Codes: B60L 2200/26 Rail vehicles
• CPC Class Codes: B61L 15/0062 On-board target speed calculation or supervision
• CPC Class Codes: Y02T 10/64 Electric machine technologies in electromobility
• CPC Class Codes: Y10S 706/90 Fuzzy logic

Abstract

Disclosed are a method and a device for stopping a vehicle at a predetermined position in which the distance between actual and target positions of the vehicle and the velocity and accleration of the vehicle are detected so as to perform brake control by selecting brake control notches on the basis of the detected values of the distance, the velocity and the acceleration to thereby stop the vehicle at the predetermined position, are featured in that the stop gap accuracy in the case where the notch presently on actuation is maintained as well as the stop gap accuracy are obtained as fuzzy values and compared with each other to thereby perform the notch selection.

Claims

1. A method for controlling a controllable subject to adapt the behavior of the subject to an optimum condition as to a predetermined control purpose, comprising:

a first step of specifying a value of at least a first signal representing a current status of said subject;

a second step of predicting, by using said specified value of said first signal, values of a parameter for evaluating a condition of said control purpose according to a predetermined formula, said values being those values which would be obtained, respectively, as a result of control operations under a plurality of different control instructions, if applied to said subject, respectively;

a third step of converting each of said values of the parameter to at least one fuzzy value determined according to at least one of membership functions with respect to preselected performance indices relating to said control purpose, each of said membership functions defining a relationship between the fuzzy value and the value of the parameter; and

fourth step of determining one control instruction to be applied to said subject among said different control instructions based on the fuzzy values of said parameter.

2. A method according to claim 1, wherein said different control instructions include at least a first instruction which is applied to said subject at present.

3. A method according to claim 2, wherein said determination of one control instruction is carried out by selecting one of predetermined rules, each of which comprises a first part designating one of said performance indices and a second part designating the control instruction to be selected, on the basis of the fuzzy values with respect to the performance indices defined in the first part of said respective rules, thereby selecting the control instruction defined in the selected rule.

4. A method for controlling a controllable subject to adapt the behavior of the subject to an optimum condition as to a predetermined control purpose, comprising:

a first step of specifying a value of least at first signal representing a current status of said subject;

a second step of predicting, by using said specified value of said first signal, values of a parameter for evaluating a condition of said control purpose according to a predetermined formula, said values being those values which would be obtained, respectively, as a result of control operations under a plurality of different control instructions, if applied to said subject, respectively;

a third step of converting each of said values of the parameter to at least one fuzzy value determined according to at least one of membership functions with respect to preselected performance indices relating to said control purpose, each of said membership functions defining relationship between the fuzzy value and the value of the parameter; and

fourth step of determining whether a control instruction applied to said subject at present should be changed or not by evaluating the fuzzy values of said parameter.

5. A method for determining a control instruction for controlling a controllable subject in an optimum condition on the basis of a set of fuzzy reasoning control rules and membership functions, each of said control rules having a first part and a second part, said first part designating a relationship between a control instruction and at least one performance index relating to a predetermined control purpose of said controllable subject operated under said control instruction, said second part indicating a control instruction to be executed, each of said membership functions defining a relationship between a fuzzy value of one of said performance indices and a predicted value of a parameter for evaluating a condition of said control purpose, said method comprising:

a step of predicting, by using at least one value representing a current status of said controllable subject, values of said parameter obtained under a plurality of different control instructions;

a step of converting said predicted values of sad parameters to fuzzy values relating to at least one of said membership functions; and

a step of determining one control instruction to be applied to said controllable subject among said different control instructions by evaluating said converted fuzzy values on the basis of said control rules.

6. A system for determining a control instruction to control a controllable subject in an optimum condition as to a predetermined control purpose, comprising:

means for storing a set of predetermined control rules, each of said control instructions comprising an if part and a then part, said if part designating an evaluation of at least one performance index relating to a control purpose of said controllable subject in combination with a control instruction which has an effect on a value of said performance index, said then part designating an execution of said control instruction;

means for predicting, by using at least one value representing a current status of said controllable subject, values of at least one parameter for evaluating a condition of said control purpose obtained under a plurality of different control instructions;

means for converting said predicted values of said parameters to fuzzy values relating to at least one of a plurality of membership functions, wherein each of said membership functions defines a relationship between a fuzzy value of one of said performance indices and a predicted value of said parameter for evaluating a condition of said control purpose; and

means for evaluating said converted fuzzy values by using a set of said control rules stored in said storing means to determine a control instruction defined in said then part of a control rule which shows the best satisfaction as to said if part among the set of control rules, whereby said determined control instruction is applied to said controllable subject.

7. A system for determining a control instruction to be applied to a controllable subject by fuzzy reasoning, comprising:

means for storing a set of predetermined control rules, each of said control instructions comprising an if part and a then part, said if part including a statement for designating an evaluation of at least one performance index relating to a control purpose of said controllable subject in combination with a control instruction which has an effect on a value of said performance index, said then part including a statement for designating an execution of said control instruction defined in said if part;

means for predicting, by using at least one value representing a current status of said controllable subject and a plurality of different control instructions, values of at least one parameter for evaluating a condition of a control purpose of said controllable subject; and

means for converting said predicted values of said parameters to fuzzy values relating to performance indices defined in said control rules by using at least one membership function which defines a relationship between a fuzzy value of one of said performance indices and a predicted value of said parameter for evaluating a condition of said control purpose; and

means for determining a control instruction to be executed by evaluating said converted fuzzy values by fuzzy reasoning on the basis of said if part of said control rules stored in said storing means.