Neural network based automatic limit prediction and avoidance system and method

• US 6,332,105 B1
• Filed: 05/22/2000
• Issued: 12/18/2001
• Est. Priority Date: 05/21/1999
• Status: Expired due to Fees

• Current Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Original Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Inventors: Horn, Joseph F., Calise, Anthony J., Prasad, Jonnalagadda V. R.
• Primary Examiner(s): Camby, Richard M.
• Application Number:
  US09/576,369
• Time in Patent Office: 575 Days
• Field of Search: 701/3, 701/4, 701/5, 701/6, 701/8, 701/9, 701/11, 701/14, 700/48, 244/175, 244/75 R
• US Class Current: 701/3
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05D 1/0825 using mathematical models

Abstract

A method for performance envelope boundary cueing for a vehicle control system comprises the steps of formulating a prediction system for a neural network and training the neural network to predict values of limited parameters as a function of current control positions and current vehicle operating conditions. The method further comprises the steps of applying the neural network to the control system of the vehicle, where the vehicle has capability for measuring current control positions and current vehicle operating conditions. The neural network generates a map of current control positions and vehicle operating conditions versus the limited parameters in a pre-determined vehicle operating condition. The method estimates critical control deflections from the current control positions required to drive the vehicle to a performance envelope boundary. Finally, the method comprises the steps of communicating the critical control deflection to the vehicle control system; and driving the vehicle control system to provide a tactile cue to an operator of the vehicle as the control positions approach the critical control deflections.

Claims

1. A vehicle performance envelope boundary cueing method for a vehicle control system, said method comprising the steps of:

formulating a prediction system for a neural network; and

training said neural network to predict values of limited parameters as a function of current control positions and current vehicle operating conditions.

2. The method of claim 1, wherein said step of formulating a neural network prediction system comprises the steps of:

establishing a functional dependence of said limited parameters in said pre-determined vehicle operating condition with said current vehicle operating conditions and said current control positions; and

determining a drive method for predicting a control deflect ion required to drive the vehicle to said limited parameter.

3. The method of claim 1, further comprising the steps of:

applying said neural network to said control system of the vehicle;

measuring conditions;

generating a map of current control positions and vehicle operating conditions versus said limited parameters in a pre-determined vehicle operating condition;

communicating to the vehicle control system a control deflections required to drive the vehicle to a performance envelope boundary; and

driving the vehicle control system to provide a tactile cue to an operator of the vehicle as the control positions approach said control deflections.

4. The method of claim 3, wherein said step of measuring conditions further comprises the steps of:

measuring current control positions; and

measuring current vehicle operating conditions.

5. The method of claim 4, wherein said step of communicating control deflection further comprises the steps of:

estimating control deflections from said current control positions required to drive the vehicle to a performance envelope boundary; and

estimating critical control deflections from said current control positions required to drive the vehicle to said performance envelope boundary.

6. The method of claim 5, wherein said step of estimating critical control deflections from said current control positions required to drive the vehicle to said vehicle performance envelope boundary, comprises the steps of:

generating a linearized representation of said map; and

generating a pseudo-inverse of said linearized representation of said map;

wherein said pseudo-inverse of said linearized representation of said map is used in the step of determining critical control deflections from said measurable control positions required to drive the vehicle to said performance envelope boundary.

7. The method of claim 3, wherein said tactile cue comprises a soft-stop imposed on a vehicle controller of said vehicle control system.

8. The method of claim 3, wherein said pre-determined operating condition comprises dynamic trim;

wherein said dynamic trim is defined by vehicle angular acceleration having a value of zero and vehicle angular rate and vehicle transitional acceleration having non-zero values.

9. The method of claim 3, wherein said vehicle control system comprises a force-feel control system.

10. The method of claim 9, wherein said force-feel control system comprises a spring-mass damper system.

11. The method of claim 1, wherein said vehicle comprises a helicopter.

12. A performance envelope boundary cueing method for a vehicle control system, said method comprising the steps of:

providing a neural network trained to predict values of limited parameters as a function of current control positions and current vehicle operating conditions;

providing condition measurements to said neural network;

generating a map of current control positions and operating conditions versus limited parameters in a pre-determined operating condition;

communicating control deflections required to drive the vehicle to a performance envelope boundary to said vehicle control system; and

driving said vehicle control system to provide a tactile cue to a controller of said vehicle control system as at least one of said control positions approaches at least one of said control deflections.

13. The method of claim 12, wherein said step of providing a neural network trained to predict values, comprises the steps of:

establishing a functional dependence of said limited parameters in said pre-determined vehicle operating condition with said current vehicle operating conditions and said current control positions;

determining a method for predicting a control deflection required to drive said limited parameter to said performance boundary envelope; and

training said neural network to predict values of limited parameters as a function of current control positions and current vehicle operating conditions.

14. The method of claim 12, wherein said step of providing condition measurements to said neural network further comprises the steps of:

measuring said current control position;

communicating said current control position to said neural network;

measuring said current vehicle operating conditions; and

communicating said current vehicle operating conditions to said neural network.

15. The method of claim 14, wherein said step of communicating control deflections further comprises the steps of:

estimating control deflections from said current control positions that will result in the vehicle reaching a performance envelope boundary; and

estimating critical control deflections from said current control positions required to drive the vehicle to said performance envelope boundary.

16. The method of claim 15, wherein said step of estimating critical control deflections from said current control positions required to drive the vehicle to said performance envelope boundary, comprises the steps of:

generating a linearized representation of said map; and

generating a pseudo-inverse of said linearized representation of said map;

wherein said pseudo-inverse of said linearized representation of said map is used in the step of determining critical control deflections from said current control positions required to drive the vehicle to said performance envelope boundary.

17. The method of claim 12, wherein said vehicle control system comprises a force-feel control system.

18. The method of claim 17, wherein said force-feel control system comprises a spring-mass damper system.

19. The method of claim 12, wherein said tactile cue comprises a soft-stop imposed on a vehicle controller of said vehicle control system.

20. The method of claim 12, wherein said pre-determined operating condition comprises dynamic trim;

wherein said dynamic trim is defined by vehicle angular acceleration having a value of zero and vehicle angular rate and vehicle transitional acceleration having non-zero values.

21. The method of claim 12, wherein said vehicle comprises a helicopter.

22. A control system apparatus for providing a tactile cue to a vehicle controller upon continued deflection of the vehicle controller will result in the vehicle approaching a vehicle performance envelope boundary of the vehicle, said apparatus comprising:

a sensor;

a neural network, said neural network being trained to generate a map of current controller positions and vehicle operating conditions versus pre-determined limited parameters in a pre-determined operating condition;

a limit margin estimator, said limit margin estimator determining controller deflections from said current controller positions that will result in the vehicle reaching a performance envelope boundary;

critical limit selection logic, said logic determining critical controller deflections from said current controller positions required to drive the vehicle to said performance envelope boundary; and

a controller being tactilely adjustable, wherein deflection of said controller is inhibited as said deflection approaches at least one of said critical control deflections from at least one of said current control positions.

23. The apparatus of claim 22, wherein said sensor further comprises:

a vehicle operating condition sensor; and

a current controller position sensor.

24. The apparatus of claim 22, wherein said control system apparatus comprises a force-feel control system.

25. The apparatus of claim 24, wherein said force-feel control system apparatus comprises a spring-mass damper system.

26. The apparatus of claim 22, wherein said pre-determined operating condition comprises dynamic trim.

27. The apparatus of claim 22, wherein said vehicle comprises a helicopter.

28. A vehicle performance envelope boundary cueing system for a vehicle control system, said system comprising the steps of:

measuring means for measuring conditions;

calculation means for generating a map of current control positions and current operating conditions versus limited parameters in a pre-determined operating condition;

communication means for communicating said critical control deflection to the control system of the vehicle; and

driving means for driving said vehicle control system to provide a tactile cue to a controller of said vehicle control system which is sensed by an operator of the vehicle as at least on of said current control positions approaches at least one of said critical control deflections.

29. The system of claim 28, wherein said measuring means further comprises:

control measuring means for measuring current control positions; and

operating measuring means for measuring current vehicle operating conditions.

30. The system of claim 29, wherein said communication means further comprises:

estimation means for estimating control deflections from said current control positions required to drive the vehicle to a performance envelope boundary; and

estimation means for estimating critical control deflections from said current control positions required to drive the vehicle to said performance envelope boundary.

31. The system of claim 28, wherein said calculation means comprises a neural network.

32. The system of claim 28, wherein said pre-determined operating condition comprises dynamic trim.

33. The system of claim 28, wherein said vehicle control system comprises a force-feel control system.

34. The system of claim 33, wherein said force-feel control system comprises a spring-mass damper system.

35. The system of claim 28, wherein said vehicle comprises a helicopter.

36. A vehicle performance envelope boundary cueing method for a vehicle control system, said method comprising the steps of:

predicting future values of a limited parameter of said vehicle as a function of a current control position;

calculating estimates of a control deflection from said current control position which will cause said limited parameter to reach a specified performance envelope boundary at a future time.

37. The method of claim 36, further comprising the step of:

driving a variable force-feel control system of said vehicle to cue an operator of the approaching performance envelope boundary before reaching said performance envelope boundary.