Method and system to detect and mitigate sensor degradation

• US 10,526,992 B2
• Filed: 04/05/2017
• Issued: 01/07/2020
• Est. Priority Date: 04/05/2017
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations LLC (General Motors Company)
• Inventors: Whitney, Christopher E., Wong, Kevin C., Gilbert, Marcus S., Hutmacher, Jeffrey M.
• Primary Examiner(s): Nguyen, Hung Q
• Assistant Examiner(s): Monahon, Brian P
• Application Number:
  US15/479,855
• Publication Number:
  US 20180291832A1
• Time in Patent Office: 1,007 Days
• CPC Class Codes: F02D 2041/286 comprising means for signal processing
• CPC Class Codes: F02D 41/1495 Detection of abnormalities in the air/fuel ratio feedback system
• CPC Class Codes: F02D 41/222 relating to the failure of sensors or parameter detection devices
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G07C 5/0808 Diagnosing performance data testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005
• CPC Class Codes: Y02T 10/40 Engine management systems

Abstract

A method to detect and mitigate sensor degradation in an automobile system includes: collecting output signal data from at least one of a sensor and an actuator which is outputting the signal data related to operational parameters of a vehicle system; placing the sensor or the actuator in communication with a fault box used to purposely corrupt the output signal data; analyzing patterns of the output signal data compared to signal data from a nominal operating sensor or actuator using an artificial intelligence program; identifying when a statistical range of the patterns exceeds a first threshold level; and modifying a control signal to change the operational parameters of the vehicle system.

Claims

1. A method to detect and mitigate sensor degradation in an automobile system, comprising:

collecting an output signal data from at least one device which is outputting the signal data related to operational parameters of a vehicle system;

analyzing patterns of the output signal data compared to signal data from a nominal operating device;

identifying when the patterns exceed a threshold level indicating the at least one device is operating in a degraded condition;

introducing corrupted data into the at least one device;

identifying a range of deviation of the corrupted data from the output signal data;

modifying a control signal to change the operational parameters of the vehicle system;

identifying when the patterns exceed a second threshold level indicating the at least one device is operating in a failed condition, and outputting a device failed signal; and

generating an anticipated failure point when the at least one device reaches a predetermined percentage of the at least one device failed signal at a measurable vehicle operating condition defining a present vehicle total mileage value.

2. The method to detect and mitigate sensor degradation in an automobile system of claim 1, further including analyzing patterns of the output signal data using an artificial intelligence program.

3. The method to detect and mitigate sensor degradation in an automobile system of claim 2, further including collecting the corrupted data and the output signal data from the at least one device in an artificial intelligence fault predictor.

4. The method to detect and mitigate sensor degradation in an automobile system of claim 1, further including generating changes to a system control signal using an artificial intelligence compensation module and applying the changes to maintain signal control of the at least one device.

5. The method to detect and mitigate sensor degradation in an automobile system of claim 1, further including placing the at least one device in communication with a fault box used to purposely corrupt the output signal data.

6. The method to detect and mitigate sensor degradation in an automobile system of claim 5, further including changing at least one setting in the fault box to simulate a degrading at least one device.

7. The method to detect and mitigate sensor degradation in an automobile system of claim 1, wherein the at least one device defines one of:

a sensor used to determine pressure, temperature, position, acceleration, chemical constituent, mass flow, voltage, or current; and

an actuator used for one of a fuel injector, a throttle blade, a turbo wastegate, a CAM phaser, a spark plug, a fuel pump, an exhaust gas recirculation device, an active fuel management device, a variable lift cam, an alternator current, an electrical current, and a variable geometry turbo.

8. A method to detect and mitigate sensor degradation in an automobile system, comprising:

collecting an output signal data from at least one of a sensor and an actuator which is outputting the output signal data related to operational parameters of a vehicle system in an artificial intelligence fault predictor;

placing the sensor or the actuator in communication with a fault box used to purposely corrupt the output signal data and introduce corrupted data into the output signal data of the at least one sensor and the actuator;

analyzing patterns of the corrupted output signal data compared to a nominal signal data from a nominal operating sensor or actuator using an artificial intelligence program;

identifying when a statistical range of deviation of the corrupted data from the output signal data exceeds a first threshold level; and

modifying a control signal to change the operational parameters of the vehicle system;

identifying multiple output signal data sets from the sensor or the actuator output signal data relating output signal data at each of different fault box settings;

generating a probability of sensor or actuator diagnostic pass for each of the different fault box settings; and

creating a probability of sensor or actuator diagnostic pass curve.

9. The method to detect and mitigate sensor degradation in an automobile system of claim 8, further including:

comparing a first output signal data set from the sensor or the actuator having a first +−

4 sigma (σ

) range to the output signal data set analyzed using the artificial intelligence pattern recognition program defining a second output signal data set having a +−

4 sigma range which is narrower than the first output signal data set; and

setting a failure threshold.

10. The method to detect and mitigate sensor degradation in an automobile system of claim 8, further including reducing variation range in the output signal data.

11. A method to detect and mitigate sensor degradation in an automobile system, comprising:

collecting an output signal data from at least one of a sensor and an actuator which is outputting the output signal data related to operational parameters of a vehicle system in an artificial intelligence fault predictor;

placing the sensor or the actuator in communication with a fault box used to purposely corrupt the output signal data and introduce corrupted data into the output signal data of the at least one sensor and the actuator;

analyzing patterns of the corrupted output signal data compared to a nominal signal data from a nominal operating sensor or actuator using an artificial intelligence program;

identifying when a statistical range of deviation of the corrupted data from the output signal data exceeds a first threshold level; and

modifying a control signal to change the operational parameters of the vehicle system;

communicating the output from the artificial intelligence fault predictor to a fault box settings module having predefined and saved conditional data for each of the sensor and the actuator used as the first threshold to generate a device faulted signal and a second threshold to generate a device failed signal; and

generating an anticipated failure point when the sensor or the actuator reaches a predetermined percentage of the sensor or actuator device failed signal at a measurable vehicle operating condition defining a present vehicle total mileage value.