MINING METHODOLOGY TO ELIMINATE INAPPROPRIATE SETTING OF ERROR CONDITIONS USING OPERATING PARAMETERS

• US 20110144853A1
• Filed: 12/15/2009
• Published: 06/16/2011
• Est. Priority Date: 12/15/2009
• Status: Abandoned Application

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations Incorporated (Motors Liquidation Co. GUC Trust)
• Inventors: Holland, Steven W., Davis, Jason T., Patankar, Ravindra, Subramania, Halasya Siva, Narla, Aru, Singh, Satnam, Felke, Tim
• Application Number:
  US12/638,592
• Publication Number:
  US 20110144853A1
• Time in Patent Office: Days
• US Class Current: 701/31
• CPC Class Codes: G06F 11/0739 in a data processing system embedded in automotive or aircraft systems
• CPC Class Codes: G06F 11/0751 Error or fault detection not based on redundancy power supply failures G06F1/30; network fault management H04L41/06

Abstract

A system and method for reducing or eliminating built-in tests and diagnostic trouble codes that are set as a result of improper parameter values. The method includes collecting field failure data that identifies diagnostic trouble codes and parameters of the system that are used to set diagnostic trouble codes. The method transforms the collected data into a format more appropriate for human analysis and pre-processes the transferred data to identify and remove information that could bias the human analysis. The method includes plotting linear and nonlinear combinations of operation parameters, performing data mining and analysis for detecting inappropriate settings of fault codes in the pre-processed data and providing the mined data to a subject matter expert for review to determine whether a diagnostic trouble code has been issued because of improper parameters.

Claims

1. A method for reducing or eliminating improper error conditions in a system, said method comprising:

collecting field failure data identifying built-in tests, diagnostic trouble codes and parameters of the system that are used to set the diagnostic trouble codes;

transforming the collected data into a format more appropriate for human analysis;

pre-processing the transformed data to identify and remove information that could bias the human analysis;

performing data mining and analysis for detecting inappropriate settings of fault codes in the pre-processed data; and

providing the mined data to a subject matter expert for review to determine whether a diagnostic trouble code has been triggered because of improper values of operating parameters.

2. The method according to claim 1 wherein the parameters for a particular diagnostic trouble code include voltage, temperature and pressure.

3. The method according to claim 1 wherein transforming the collected data for analysis includes converting a chart that includes lines of information identifying diagnostic trouble codes, parameter identifier descriptions and parameter identifier values and units to a user friendly format.

4. The method according to claim 1 wherein pre-processing the transformed data includes identifying missing values and information.

5. The method according to claim 1 wherein performing data mining and analysis includes employing decision trees to reconstruct a probabilistic fault tree using parameter identifiers.

6. The method according to claim 1 wherein performing data mining and analysis includes providing plots that identify relationships between parameters and between diagnostic trouble codes and parameters.

7. The method according to claim 1 wherein providing the mined data to subject matter expert includes determining whether a particular parameter could exist for a particular system condition.

8. The method according to claim 1 further comprising data post-processing and visualization of the mined data before the data is provided to the subject matter expert.

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

10. A method for reducing or eliminating improper error conditions in a vehicle, said method comprising:

collecting field failure data identifying diagnostic trouble codes and parameters of the vehicles that are used to set the diagnostic trouble codes, said parameters including one or more of voltage, pressure and temperature;

transforming the collected data into a format more appropriate for human analysis;

pre-processing the transformed data to identify and remove information that could bias the human analysis including identifying missing values and information;

performing data mining and analysis for detecting inappropriate settings of fault codes in the pre-processed data that includes employing decision trees to reconstruct a probabilistic fault tree using parameter identifiers;

providing the mined data to a subject matter expert for review to determine whether a diagnostic trouble code has been triggered because of improper values of operating parameters that includes determining whether a particular parameter could exist for a particular system condition; and

providing recommendations to adjust the setting of the diagnostic trouble code if it is determined that a particular parameter could not exist for a particular system condition.

11. The method according to claim 10 wherein transforming the collected data for analysis includes converting a chart that includes lines of information identifying diagnostic trouble codes, parameter identifiers, parameter identifier descriptions and parameter identifier values and units to a user friendly format.

12. The method according to claim 10 further comprising data post-processing and visualization of the mined data before the data is provided to the subject matter expert.

13. The method according to claim 10 wherein transforming the collected data for analysis includes converting a chart that includes lines of information identifying diagnostic trouble codes, parameter identifiers, parameter identifier descriptions and parameter identifier values and units to a user friendly format.

14. A method for reducing or eliminating improper error signals in a system, said method comprising:

collecting data identifying built-in tests, trouble codes and parameters of the system that are used to set the trouble codes;

transforming the collected data into a format more appropriate for human analysis;

performing data mining and analysis for detecting inappropriate settings of fault codes in the pre-processed data; and

providing the mined data to a subject matter expert for review to determine whether a diagnostic trouble code has been triggered because of improper values of operating parameters.

15. The method according to claim 14 further comprising pre-processing the transformed data to identify and remove information that could bias the human analysis including identifying missing values and information.

16. The method according to claim 14 wherein the parameters for a particular trouble code include voltage, temperature and pressure.

17. The method according to claim 14 wherein performing data mining and analysis includes employing decision trees to reconstruct a probabilistic fault tree using parameter identifiers.

18. The method according to claim 14 wherein performing data mining and analysis includes providing plots that identify relationships between parameters.

19. The method according to claim 14 wherein providing the mined data to a subject matter expert includes determining whether a particular parameter could exist for a particular system condition.

20. The method according to claim 14 further comprising providing data post-processing and visualization of the mined data before the data is provided to the subject matter expert.