Ranking fault conditions

Patent metadata

Field	Value
Publication number	US-11715338-B2
Application number	US-202117146714-A
Country	US
Kind code	B2
Filing date	Jan 12, 2021
Priority date	Jan 12, 2021
Publication date	Aug 1, 2023
Grant date	Aug 1, 2023

Abstract

A plurality of fault conditions are detected on a communication network onboard a vehicle. The detected fault conditions, a fault condition importance, environment conditions, and a vehicle operation mode are input to a neural network that outputs rankings for respective detected fault conditions. The neural network is trained by determining a loss function based on a maximum likelihood principle that determines a probability distribution that ranks the detected fault conditions. The vehicle is operated based on the rankings of the fault conditions.

First claim

What is claimed is:

1. A system, comprising a computer including a processor and a memory, the memory storing instructions executable by the processor to:
detect, on a communication network onboard a vehicle, a plurality of fault conditions;
input the detected fault conditions, a fault condition importance, environment conditions, and a vehicle operation mode to a neural network that outputs rankings for respective detected fault conditions, wherein the neural network is trained by determining a loss function based on a maximum likelihood principle that determines a probability distribution that ranks the detected fault conditions; and
operate the vehicle based on the rankings of the fault conditions.

2. The system of claim 1 , wherein environment conditions include at least one of road data or weather data.

3. The system of claim 1 , wherein the fault condition importance is determined based on an amount of time the vehicle is permitted to operate after detecting corresponding fault conditions.

4. The system of claim 1 , wherein the neural network includes first hidden layers that output latent variables to second hidden layers that output rankings for the detected fault conditions.

5. The system of claim 4 , wherein the first hidden layers include bias neurons.

6. The system of claim 1 , wherein the neural network optimizes parameters of the loss function by applying gradient descent to the loss function.

7. The system of claim 1 , wherein the instructions further include instructions to operate the vehicle based further on identifying a highest ranked fault condition as one of persistent or transient.

8. The system of claim 1 , wherein the instructions further include instructions to, upon determining an amount of time between detecting the fault conditions and receiving the output from the neural network is greater than or equal to a threshold, monitor the communications network onboard the vehicle for updated fault conditions and maintain operation of the vehicle.

9. The system of claim 8 , wherein the instructions further include instructions to, upon detecting updated fault conditions, input the updated fault conditions into the neural network that outputs a ranking for the updated fault condition.

10. The system of claim 1 , wherein the instructions further include instructions to operate the vehicle based further on determining an amount of time between detecting the fault conditions and receiving the output from the neural network is less than a threshold.

11. The system of claim 1 , wherein instructions further include instructions to determine the vehicle operation mode based on vehicle sensor data.

12. The system of claim 1 , wherein the instructions further include instructions to determine the environment conditions based on vehicle sensor data.

13. The system of claim 1 , wherein the instructions further include instructions to access a look-up table to determine the fault condition importance.

14. The system of claim 1 , wherein the instructions further include instructions to, based on identifying a highest ranked fault condition, one of maintain operation of the vehicle or perform a minimal risk maneuver.

15. A method, comprising:
detecting, on a communication network onboard a vehicle, a plurality of fault conditions;
inputting the detected fault conditions, a fault condition importance, environment conditions, and a vehicle operation mode to a neural network that outputs rankings for respective detected fault conditions, wherein the neural network is trained by determining a loss function based on a maximum likelihood principle that determines a probability distribution that ranks the detected fault conditions; and
operating the vehicle based on the rankings of the fault conditions.

16. The method of claim 15 , wherein the neural network optimizes parameters of the loss function by applying gradient descent to the loss function.

17. The method of claim 15 , further comprising operating the vehicle based further on determining an amount of time between detecting the fault conditions and receiving the output from the neural network is less than a threshold.

18. The method of claim 15 , further comprising, based on identifying a highest ranked fault condition, one of maintaining operation of the vehicle or performing a minimal risk maneuver.

19. The method of claim 15 , further comprising operating the vehicle based further on identifying a highest ranked fault condition as one of persistent or transient.

20. The method of claim 15 , wherein the fault condition importance is determined based on an amount of time the vehicle is permitted to operate after detecting corresponding fault conditions.

Assignees

• Ford Global Tech Llc

Inventors

• Vejalla Dhanunjay
• Symanow David A
• Siciak Ray

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G07C5/0816 Primary

  Indicating performance data, e.g. occurrence of a malfunction · CPC title

• G01C21/3691

  Retrieval, searching and output of information related to real-time traffic, weather, or environmental conditions (arrangements for giving variable traffic instructions G08G1/09) · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title