Enhanced vehicle operation

Patent metadata

Field	Value
Publication number	US-11801836-B2
Application number	US-202017082507-A
Country	US
Kind code	B2
Filing date	Oct 28, 2020
Priority date	Oct 28, 2020
Publication date	Oct 31, 2023
Grant date	Oct 31, 2023

Abstract

A computer includes a processor and a memory storing instructions executable by the processor to collect operation data of one or more components of a vehicle, input the collected operation data and previously stored operation data to a first machine learning program that assigns the collected operation data to one of a plurality of previously determined paths and outputs a target fuel consumption rate of the vehicle based on the assigned path, input the target fuel consumption rate and the collected operation data to a second machine learning program that outputs a plurality of operation settings of the one or more components to attain the target fuel consumption rate, and actuate the one or more components to attain the plurality of operation settings.

First claim

The invention claimed is:

1. A system, comprising a computer including a processor and a memory, the memory storing instructions executable by the processor to:
collect operation data of one or more components of a vehicle;
input to a first machine learning program (a) a current trip vector that includes a set of values including a predicted path of the vehicle and one or more current operation settings for the components, (b) the collected operation data, and (c) previously stored operation data;
wherein the first machine learning program assigns the collected operation data to one of a plurality of previously determined trip vectors that include one of a plurality of previously determined paths,
assigns the current trip vector to one of a plurality of clusters, wherein each of the plurality of clusters is based on a respective previously determined trip vector including respective predicted paths and respective one or more current operation settings for the components, to determine a predicted trip vector including a further predicted path and further one or more current operation settings for the components, and
outputs a target fuel consumption rate of the vehicle based on the predicted trip vector and the previously determined trip vectors;

input the target fuel consumption rate and the collected operation data to a second machine learning program that outputs a plurality of operation settings of the one or more components to attain the target fuel consumption rate; and
actuate the one or more components to attain the plurality of operation settings.

2. The system of claim 1 , wherein the instructions further include instructions to collect the operation data from a predetermined period of time from activation of the vehicle.

3. The system of claim 1 , wherein the instructions further include instructions to identify an occupant of the vehicle based on the collected operation data.

4. The system of claim 3 , wherein the instructions further include instructions to input the identification of an occupant of the vehicle to the first machine learning program and to output the target fuel consumption rate based on the identification of the occupant.

5. The system of claim 4 , wherein the instructions further include instructions to input previously determined operation data of the vehicle operated by the identified occupant to the first machine learning program.

6. The system of claim 1 , wherein the instructions further include instructions to input a planned path of the vehicle to the first machine learning program to output the target fuel consumption rate.

7. The system of claim 1 , wherein each previously determined trip vector defining the plurality of clusters includes an origin, a destination, and a path between the origin and the destination, and each previously determined trip vector has at least one of a different destination or a different path from each other previously determined trip vector.

8. The system of claim 1 , wherein the instructions further include instructions to input the collected operation data and previously stored operation data to a third machine learning program trained to output the trip vector.

9. The system of claim 1 , wherein the plurality of operation settings includes a prescribed torque output from a powertrain, and the instructions further include instructions to actuate the powertrain to attain the prescribed torque output.

10. The system of claim 1 , wherein the plurality of operation settings includes a prescribed maximum speed, and the instructions further include instruction to actuate at least one of a powertrain, a propulsion, or a brake to operate the vehicle below the prescribed maximum speed.

11. The system of claim 1 , wherein the instructions further include instructions to input a current timestamp to the first machine learning program and to output the target fuel consumption rate based on the current timestamp.

12. A method, comprising:
collecting operation data of one or more components of a vehicle;
inputting to a first machine learning program (a) a current trip vector that includes a set of values including a predicted path of the vehicle and one or more current operation settings for the components, (b) the collected operation data, and (c) previously stored operation data;
wherein the first machine learning program assigns the collected operation data to one of a plurality of previously determined trip vectors that include one of a plurality of previously determined paths,
assigns the current trip vector to one of a plurality of clusters, wherein each of the plurality of clusters is based on a respective previously determined trip vector including respective predicted paths and respective one or more current operation settings for the components, to determine a predicted trip vector including a further predicted path and further one or more current operation settings for the components, and
outputs a target fuel consumption rate of the vehicle based on the predicted trip vector and the previously determined trip vectors;

inputting the target fuel consumption rate and the collected operation data to a second machine learning program that outputs a plurality of operation settings of the one or more components to attain the target fuel consumption rate; and
actuating the one or more components to attain the plurality of operation settings.

13. The method of claim 12 , further comprising collecting the operation data from a predetermined period of time from activation of the vehicle.

14. The method of claim 12 , further comprising identifying an occupant of the vehicle based on the collected operation data.

15. The method of claim 14 , further comprising inputting the identification of an occupant of the vehicle to the first machine learning program and to output the target fuel consumption rate based on the identification of the occupant.

16. The method of claim 12 , further comprising inputting a planned path of the vehicle to the first machine learning program to output the target fuel consumption rate.

17. The method of claim 12 , wherein each previously determined trip vector defining the plurality of clusters includes an origin, a destination, and a path between the origin and the destination, and each previously determined trip vector has at least one of a different destination or a different path from each other previously determined trip vector.

18. The method of claim 12 , wherein the plurality of operation settings includes a prescribed torque output from a powertrain, and the method further comprises actuating the powertrain to attain the prescribed torque output.

Assignees

• Ford Global Tech Llc

Inventors

• Wang Lijun
• Zhang Chen
• Kreucher Rebecca

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• B60W30/18 Primary

  Propelling the vehicle · CPC title

• B60W10/06

  including control of combustion engines · CPC title

• B60W40/08 Primary

  related to drivers or passengers · CPC title