HEALTH PROGNOSIS FOR COMPLEX SYSTEM USING FAULT MODELING

• US 20110202227A1
• Filed: 02/17/2010
• Published: 08/18/2011
• Est. Priority Date: 02/17/2010
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations Incorporated (Motors Liquidation Co. GUC Trust)
• Inventors: Howell, Mark N., Bae, Hong S., Zhang, Yilu, Shin, Kwang-Keun, Salman, Mutasim A., Rajagopalan, Satish, Tang, Xidong
• Granted Patent: US 8,260,493 B2
• Time in Patent Office: Days
• US Class Current: 701/29
• CPC Class Codes: B60W 2510/244 Charge state
• CPC Class Codes: B60W 2510/248 Age of storage means
• CPC Class Codes: B60W 50/0097 Predicting future conditions
• CPC Class Codes: B60W 50/04 Monitoring the functioning of the control system

Abstract

A system and method for providing component and sub-system state of health prognosis in a complex system using fault models and component aging models. The method includes determining a current state of health value for a sub-system using fault signature test results and determining current state of health values for a plurality of components in the sub-system using the fault signature test results. The method also determines current state of health values for components in the system that cannot use fault signature test results using a first probability model and the current state of health values for the plurality of components. The method determines predicted future state of health values for the components in the sub-system using component aging models and determines a predicted future state of health value for the sub-system using a second probability model and the future state of health values of the components.

Claims

1. A method for determining the state of health of a sub-system in a system, said method comprising:

collecting fault signature test results;

determining a current state of health value for the sub-system using the fault signature test results;

determining current state of health values of a plurality of components in the sub-system using the fault signature test results;

determining current state of health values for components in the system that cannot use fault signature test results using a first probability model and the current state of health value for the sub-system and the current state of health values of the components determined by the fault signature test results;

determining predicted future state of health values for the components in the sub-system using component aging models;

determining a predicted future state of health values for the sub-system using a second probability model and the future state of health value for the components;

determining if the predicted state of health value of the sub-system is greater than a threshold; and

determining the remaining useful life of the sub-system if the predicted state of health value of the sub-system is less than the threshold.

2. The method according to claim 1 further comprising advancing a future time period and determining predicted future state of health values for the components in the sub-system and determining a predicted state of health value for the sub-system if the predicted state of health value of the sub-system at the future time is less than the threshold.

3. The method according to claim 1 wherein determining a predicted future period state of health value for the sub-system using a second probability model includes using a fault model selected from the group consisting of Bayesian network models, dynamic Bayesian network models, hidden Markov models, fuzzy logic models, belief network models and Petri net models.

4. The method according to claim 1 wherein the aging model is selected from the group consisting of Arrhenius equation aging models and Paris equation aging models.

5. The method according to claim 1 wherein determining predicted state of health values using aging models includes using the equation:

SOH_C(τ

)=g[SOH_C(τ

−

1),u]

where SOH is state of health, u is an environmental variable and τ

is a time period.

6. The method according to claim 1 wherein determining current state of health values of a plurality of components using fault signature test results includes using fault signature test results based on sensor readings, measurements, simulations and expert knowledge.

7. The method according to claim 1 wherein the system is a vehicle system.

8. The method according to claim 7 wherein the vehicle system is a vehicle electrical system.

9. A method for determining the state of health of a sub-system in a system, said method comprising:

determining predicted future state of health values for a plurality of components in the sub-system using component aging models;

determining a predicted future state of health value for the sub-system using a probability model and the predicted future state of health values;

determining if the predicted state of health value for the sub-system is greater than a threshold; and

determining the remaining useful life of the sub-system if the predicted state of health value for the sub-system is less than the threshold.

10. The method according to claim 9 further comprising advancing a future time period and determining predicted future state of health values of the components in the sub-system and determining a predicted state of health value for the sub-system if the predicted state of health value of the sub-system at the future time is less than the threshold.

11. The method according to claim 9 wherein determining a predicted future period state of health value for the sub-system using a probability model includes using a fault model selected from the group consisting of Bayesian network models, dynamic Bayesian network models, hidden Markov models, fuzzy logic models, belief network models and Petri net models.

12. The method according to claim 9 wherein determining predicted state of health values using aging models includes using the equation:

SOH_C(τ

)=g[SOH_C(τ

−

1),u]

where SOH is state of health, u is an environmental variable and τ

is a time period.

13. The method according to claim 9 wherein the aging model is selected from the group consisting of Arrhenius equation aging models and Paris equation aging models.

14. A system for determining the state of health of a sub-system in a complex system, said system comprising:

means for collecting fault signature test results;

means for determining a current state of health value for the sub-system using the fault signature test results;

means for determining current state of health values for a plurality of components in the sub-system using the fault signature test results;

means for determining current state of health of values for components in the system than cannot use fault signature tests using a first probability model and the state of health values of the plurality of components;

means for predicting future state of health values of the components in the sub-system using component aging models;

means for determining a predicted future state of health values for the sub-system using a second probability model and the future state of health values of the components;

means for determining if the predicted state of health value for the sub-system is greater than a threshold; and

means for determining the remaining useful life of the sub-system if the predicted state of health value of the sub-system is less than the threshold.

15. The system according to claim 14 further comprising means for advancing a future time period and determining predicted future state of health values of the components in the sub-system and determining a predicted state of health value for the sub-system if the predicted state of health value of the sub-system at the future time is less than the threshold.

16. The system according to claim 14 wherein the means for determining a predicted future period state of health value for the sub-system uses a probability model includes using a fault model selected from the group consisting of Bayesian network models, dynamic Bayesian network models, hidden Markov models, fuzzy logic models, belief network models and Petri net models.

17. The system according to claim 14 wherein the means for determining predicted state of health values using aging models includes uses the equation:

SOH_C(τ

)=g[SOH_C(τ

−

1),u]

where SOH is state of health, u is an environmental variable and τ

is a time period.

18. The system according to claim 14 wherein the aging model is selected from the group consisting of Arrhenius equation aging models and Paris equation aging models.