Regenerative braking control system

Patent metadata

Field	Value
Publication number	US-11235665-B2
Application number	US-201916365162-A
Country	US
Kind code	B2
Filing date	Mar 26, 2019
Priority date	Mar 26, 2019
Publication date	Feb 1, 2022
Grant date	Feb 1, 2022

Abstract

A vehicle includes an electric machine and a controller. The electric machine is configured to draw energy from a battery to propel the vehicle and to recharge the battery during regenerative braking. The controller is programmed to, in response to identifying a regenerative braking opportunity along an upcoming road segment based on a classification of driver behavior and a classification of the upcoming road segment, operate the electric machine to recharge the battery along the upcoming road segment.

First claim

What is claimed is:

1. A vehicle comprising:
an electric machine configured to draw energy from a battery to propel the vehicle and to recharge the battery during regenerative braking; and
a controller programmed to, in response to identifying a regenerative braking opportunity along an upcoming road segment based on a classification of a current driver behavior and a classification of the upcoming road segment, operate the electric machine to recharge the battery along the upcoming road segment, wherein the classification of the upcoming road segment is selected from one of multiple road segment classifications that were established based on an application of a k-means clustering algorithm to road segment characteristic data from multiple road segments.

2. The vehicle of claim 1 , wherein the controller is further programed to, in response to a battery charge capacity being less than a value required to recoup an estimated potential regenerative braking energy along the upcoming road segment, operate the electric machine to increase the battery charge capacity to greater than the value prior to the vehicle reaching the upcoming road segment.

3. The vehicle of claim 2 , wherein data from multiple drivers and multiple road segments is utilized to determine the classification of the current driver behavior and the classification of the upcoming road segment, and wherein the estimated potential regenerative braking energy along the upcoming road segment is based on applying a nearest neighbor algorithm to the data.

4. The vehicle of claim 1 , wherein the classification of the current driver behavior is selected from one of multiple driver classifications that were established based on an application of a k-means clustering algorithm to driver behavior data from multiple drivers.

5. The vehicle of claim 4 , wherein a neural network is trained to the multiple classifications of drivers, and wherein the classification of the current driver behavior is selected from one of the multiple driver classifications based on application of the neural network to characteristics of the current driver behavior.

6. The vehicle of claim 1 , wherein a neural network is trained to the multiple road segment classifications, and wherein the classification of the upcoming road segment is selected from one of the multiple road segment classifications based on application of the neural network to characteristics of the upcoming road segment.

7. A vehicle battery recharging method:
collecting driver behavior and road segment characteristic data from multiple drivers over multiple segments;
establishing a current driver behavior classification based on the driver behavior data, wherein the current driver behavior classification is selected from one of multiple driver classifications that were established based on an application of a k-means clustering algorithm to the driver behavior data from multiple drivers;
establishing an upcoming road segment classification based on the road segment characteristic data;
identifying a regenerative braking opportunity along the upcoming road segment by cross-referencing the driver behavior classification to the upcoming road segment classification; and
recharging the battery via regenerative braking along the upcoming road segment.

8. The method of claim 7 further comprising:
in response to a battery charge capacity being less than a value required to recoup an estimated potential regenerative braking energy along the upcoming road segment, operating an electric machine to increase the battery charge capacity to greater than the value prior to the vehicle reaching the upcoming road segment.

9. The method of claim 8 , wherein the estimated potential regenerative braking energy along the upcoming road segment is based on applying a nearest neighbor algorithm to the data.

10. The method of claim 7 , wherein a neural network is trained to the multiple classifications of drivers, and wherein the current driver behavior classification is selected from one of the multiple driver classifications based on application of the neural network to characteristics of the current driver behavior.

11. The method of claim 7 , wherein the upcoming road segment classification is selected from one of multiple road segment classifications that were established based on an application of a k-means clustering algorithm to the road segment characteristic data from multiple road segments.

12. The method of claim 11 , wherein a neural network is trained to the multiple road segment classifications, and wherein the classification of the upcoming road segment is selected from one of the multiple road segment classifications based on application of the neural network to characteristics of the upcoming road segment.

13. A vehicle comprising:
an electric machine configured to draw energy from a battery to propel the vehicle and to recharge the battery during regenerative braking; and
a controller programmed to,
in response to identifying a regenerative braking opportunity along an upcoming road segment based on classifications of current driver behavior and the upcoming road segment, schedule a regenerative braking event along the upcoming road segment, wherein data from multiple drivers and multiple road segments is utilized to determine the classifications of the current driver behavior and the upcoming road segment, and
in response to a battery charge capacity being less than a value required to recoup an estimated potential regenerative braking energy during the scheduled regenerative braking event, operate the electric machine to increase the battery charge capacity to greater than the value prior to the scheduled regenerative braking event, wherein the estimated potential regenerative braking energy during the scheduled regenerative braking event is based on applying a nearest neighbor algorithm to the data.

14. The vehicle of claim 13 , wherein the classification of the current driver behavior is selected from one of multiple driver classifications that were established based on an application of a k-means clustering algorithm to driver behavior data from multiple drivers.

15. The vehicle of claim 14 , wherein a neural network is trained to the multiple classifications of drivers, and wherein the classification of the current driver behavior is selected from one of the multiple driver classifications based on application of the neural network to characteristics of the current driver behavior.

16. The vehicle of claim 13 , wherein the classification of the upcoming road segment is selected from one of multiple road segment classifications that were established based on an application of a k-means clustering algorithm to road segment characteristic data from multiple road segments.

17. The vehicle of claim 16 , wherein a neural network is trained to the multiple road segment classifications, and wherein the classification of the upcoming road segment is selected from one of the multiple road segment classifications based on application of the neural network to characteristics of the upcoming road segment.

Assignees

• Ford Global Tech Llc

Inventors

• Mendez Pineda Herbert David
• Ostrowski David
• John Mohan
• Schein Victoria Leigh
• Herbert William G
• Santiago Estrada Madeleine Itzel

Classifications

• G06N3/09

  Supervised learning · CPC title

• B60W2050/0043

  Signal treatments, identification of variables or parameters, parameter estimation or state estimation · CPC title

• B60L53/00

  Methods of charging batteries, specially adapted for electric vehicles; Charging stations or on-board charging equipment therefor; Exchange of energy storage elements in electric vehicles · CPC title

• B60L7/18 Primary

  Controlling the braking effect (B60L7/12, B60L7/14, B60L7/16 take precedence) · CPC title

• Y02T90/16

  Information or communication technologies improving the operation of electric vehicles · CPC title