Method and apparatus for on-board/off-board fault detection

• US 8,972,099 B2
• Filed: 02/05/2013
• Issued: 03/03/2015
• Est. Priority Date: 02/05/2013
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations LLC (General Motors Company)
• Inventors: Ghoneim, Youssef A.
• Primary Examiner(s): Cheung, Mary
• Assistant Examiner(s): MAZZARA, ANNE MARIE
• Application Number:
  US13/759,389
• Publication Number:
  US 20140222281A1
• Time in Patent Office: 756 Days
• Field of Search: 701/29.1, 701/29.6, 701/29.7, 701/31.4, 701/31.6, 701/31.7, 701/32.7
• US Class Current: 701/31.4
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G06F 17/00 Digital computing or data processing equipment or methods, specially adapted for specific functions information retrieval, database structures or file system structures therefor G06F16/00

Abstract

A method for monitoring a subsystem of a subject vehicle includes, in an off-board environment, executing a controller-based scheme. The scheme includes determining a plurality of residuals based upon expected and observed states of parameters associated with operation of the subsystem, determining a subsystem operating signature based upon the residuals, employing the subsystem operating signature to isolate a subsystem fault, and communicating the subsystem fault to an on-board controller of the subject vehicle.

Claims

1. A method for monitoring a subsystem of a subject vehicle, comprising:

when an on-board algorithm detects unexpected behavior in the subsystem, communicating subsystem data to an off-board environment;

in the off-board environment, executing a controller-based scheme comprising;

determining a plurality of residuals based upon expected and observed states of parameters associated with operation of the subsystem;

determining a subsystem operating signature based upon the residuals;

employing the subsystem operating signature to isolate a subsystem fault by comparing the subsystem operating signature to a subsystem fault isolation matrix determined in an off-line control scheme; and

communicating the subsystem fault from the off-board environment to an on-board controller of the subject vehicle.

2. The method of claim 1, wherein determining the plurality of residuals comprises:

developing a plurality of residual equations associated with the parameters; and

employing the residual equations to determine the plurality of residuals based upon the expected and observed states of the parameters.

3. The method of claim 2, wherein developing the plurality of residual equations comprises developing the plurality of residual equations associated with the parameters in an off-line environment.

4. The method of claim 1, wherein determining a subsystem operating signature comprises calculating an average for each of the residuals based upon the expected and observed states of the parameters associated with operation of the subsystem over a period of time.

5. The method of claim 1, further comprising determining a fault signature associated with a candidate fault, comprising:

developing a plurality of residual equations based upon expected and observed states of parameters associated with operation of the subsystem;

inducing the candidate fault in the subsystem and observing states of the parameters;

employing the residual equations to determine said plurality of residuals;

determining average residuals over a period of time; and

determining the fault signature associated with the candidate fault based upon the average residuals.

6. The method of claim 1, wherein employing the subsystem operating signature to isolate a subsystem fault comprises:

employing elements of the subsystem operating signature to determine a plurality of fault probabilities associated with a plurality of candidate faults;

determining a probability of no fault based upon the plurality of fault probabilities associated with the candidate faults; and

isolating one of the plurality of candidate faults and no fault as the subsystem fault based upon the plurality of fault probabilities associated with the candidate faults and the probability of no fault.

7. The method of claim 6, wherein isolating one of the plurality of candidate faults and no fault comprises isolating as the subsystem fault the one of the candidate faults having a fault probability that is greater than the sum of the fault probabilities of the other candidate faults and the probability of no fault.

8. The method of claim 6, wherein isolating one of the plurality of candidate faults and no fault comprises isolating as the subsystem fault the no fault condition when the probability of no fault is greater than the sum of the fault probabilities of the candidate faults.

9. A method for monitoring a subsystem of a subject vehicle, comprising:

in an off-line environment, executing a controller-based scheme including developing a plurality of residual equations associated with the subsystem;

observing parameters of the subsystem on-vehicle and when an on-board algorithm detects unexpected behavior in the subsystem, communicating observed states of the parameters to an off-board system; and

executing a controller-based scheme in the off-board system, including;

employing the residual equations to determine a plurality of residuals based upon expected states and the observed states of the parameters;

determining a subsystem operating signature based upon the residuals;

employing the subsystem operating signature to isolate a subsystem fault by comparing the subsystem operating signature to a subsystem fault isolation matrix determined in the off-line environment; and

communicating the subsystem fault from the off-board system to the subject vehicle.

10. The method of claim 9, wherein determining a subsystem operating signature comprises calculating an average for each of the residuals based upon the expected and observed states of the parameters associated with operation of the subsystem over a period of time.

11. The method of claim 9, further comprising determining a fault signature associated with a candidate fault, including:

developing a plurality of residual equations based upon expected and observed states of parameters associated with operation of the subsystem;

inducing the candidate fault in the subsystem and observing states of the parameters;

employing the residual equations to determine said plurality of residuals;

determining average residuals over a period of time; and

determining the fault signature associated with the candidate fault based upon the average residuals.

12. The method of claim 9, wherein employing the subsystem operating signature to isolate a subsystem fault comprises:

employing elements of the subsystem operating signature to determine a plurality of fault probabilities associated with candidate faults;

determining a probability of no fault based upon the plurality of fault probabilities associated with the candidate faults; and

isolating a subsystem fault based upon the plurality of fault probabilities associated with the candidate faults and the probability of no fault.

13. The method of claim 9, wherein employing the subsystem operating signature to isolate a subsystem fault comprises:

employing elements of the subsystem operating signature to determine a plurality of fault probabilities associated with a plurality of candidate faults;

determining a probability of no fault based upon the plurality of fault probabilities associated with the candidate faults; and

isolating one of the plurality of candidate faults and no fault as the subsystem fault based upon the plurality of fault probabilities associated with the candidate faults and the probability of no fault.

14. The method of claim 13, wherein isolating one of the plurality of candidate faults and no fault comprises isolating as the subsystem fault the one of the candidate faults having a fault probability that is greater than the sum of the fault probabilities of the other candidate faults and the probability of no fault.

15. The method of claim 13, wherein isolating one of the plurality of candidate faults and no fault comprises isolating as the subsystem fault the no fault condition when the probability of no fault is greater than the sum of the fault probabilities of the candidate faults.

16. A method for monitoring a subsystem of a subject vehicle, comprising:

in an off-line environment, executing a controller-based scheme including developing a plurality of residual equations associated with the subsystem;

communicating subsystem parameters to an off-board system when an on-board algorithm detects unexpected behavior in the subsystem; and

executing a controller-based scheme in the off-board system, including;

determining a plurality of residuals based upon expected states and observed states of the subsystem parameters;

determining a subsystem operating signature based upon the residuals;

employing the subsystem operating signature to isolate a subsystem fault by comparing the subsystem operating signature to a subsystem fault isolation matrix determined in the off-line environment; and

communicating the subsystem fault from the off-board system to the subject vehicle.