System for estimating vehicle states for rollover reduction

• US 20080281482A1
• Filed: 05/07/2007
• Published: 11/13/2008
• Est. Priority Date: 05/07/2007
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations Incorporated (Motors Liquidation Co. GUC Trust)
• Inventors: Huang, Jihua
• Granted Patent: US 7,848,864 B2
• Time in Patent Office: Days
• US Class Current: 701/29
• CPC Class Codes: B60W 2030/043 about the roll axis
• CPC Class Codes: B60W 2040/1315 Location of the centre of gravity
• CPC Class Codes: B60W 2040/133 about the roll axis
• CPC Class Codes: B60W 2050/0013 Optimal controllers
• CPC Class Codes: B60W 2510/20 Steering systems
• CPC Class Codes: B60W 2520/12 Lateral speed
• CPC Class Codes: B60W 2520/14 Yaw
• CPC Class Codes: B60W 2520/18 Roll
• CPC Class Codes: B60W 2540/18 Steering angle
• CPC Class Codes: B60W 40/11 Pitch movement
• CPC Class Codes: B60W 40/112 Roll movement
• CPC Class Codes: B60W 40/114 Yaw movement
• CPC Class Codes: B60W 40/13 Load or weight

Abstract

A system and method for estimating vehicle states, such as vehicle roll-rate, vehicle roll angle, vehicle lateral velocity and vehicle yaw-rate, for use in rollover reduction. The system includes an extended Kalman filter observer responsive to a steering angle signal, a yaw-rate signal, a roll-rate signal, a speed signal and a lateral acceleration signal that calculates an estimated yaw-rate signal, an estimated roll-rate, an estimated roll angle and an estimated lateral velocity. The system also includes a lateral velocity estimation processor responsive to the roll-rate signal, the estimated roll angle signal, the estimated lateral velocity signal and the lateral acceleration signal that calculates a modified lateral velocity estimation signal when the vehicle is operating in a non-linear region.

Claims

1. A system for estimating vehicle states of yaw-rate, roll-rate, roll angle and lateral velocity, said system comprising:

a steering angle sensor providing a steering angle signal of a steering angle of the vehicle;

a yaw-rate sensor for providing a yaw-rate signal of the yaw-rate of the vehicle;

a roll-rate sensor for providing a roll-rate signal of the roll-rate of the vehicle;

a speed sensor for providing a speed signal of the speed of the vehicle;

a lateral acceleration sensor for providing a lateral acceleration signal of the lateral acceleration of the vehicle;

a Kalman filter observer responsive to the steering angle signal, the yaw-rate signal, the roll-rate signal, the speed signal and the lateral acceleration signal, said observer using Kalman filtering to provide an estimated yaw-rate signal, an estimated roll-rate, an estimated roll angle and an estimated lateral acceleration; and

a lateral velocity estimation processor responsive to the roll-rate signal, the estimated roll angle signal, the estimated lateral velocity signal and the lateral acceleration signal, said lateral velocity estimation processor calculating a modified lateral velocity estimation signal when the vehicle is operating in a non-linear region.

2. The system according to claim 1 wherein the Kalman filter observer changes a covariance matrix in the filter when the vehicle goes from a linear to a non-linear operating region.

3. The system according to claim 1 wherein the Kalman filter observer uses a three degree-of-freedom model.

4. The system according to claim 1 further comprising a sensor pre-filtering processor responsive to the steering angle signal, the yaw-rate signal, the roll-rate signal, the speed signal and the lateral acceleration signal, said sensor pre-filtering processor filtering out a low-frequency component of the signals, and estimating a sensor bias based on the low-frequency component.

5. The system according to claim 4 wherein the sensor pre-filtering processor uses a recursive least squares technique to provide the sensor bias estimation.

6. The system according to claim 4 wherein the sensor pre-filtering processor employs a variable forgetting factor that is a value between 0 and 1 and is set based on an error between the low-frequency component of the sensor signal and an estimated bias of the sensor.

7. The system according to claim 1 wherein the lateral velocity estimation processor integrates the lateral velocity to modify the lateral velocity estimation signal.

8. The system according to claim 7 wherein the lateral velocity estimation processor uses the following equation to determine the modified estimated lateral acceleration signal:

v_y=∫

(a_ym−

γ

v_x−

g sin φ

)dt

where v_y is the lateral velocity, a_ym is the lateral acceleration, γ

is the yaw-rate, v_x is the vehicle speed and g is the gravitational constant.

9. The system according to claim 1 wherein the system is used to estimate vehicle states for vehicle rollover reduction.

10. The system according to claim 1 wherein the system uses the lateral acceleration signal, the roll-rate signal or the roll angle signal to determine whether the vehicle is in the non-linear operating region.

13. The system according to claim 11 wherein the Kalman filter observer uses a three degree-of-freedom model.

14. The system according to claim 1 further comprising a sensor pre-filtering processor responsive to the sensor signals, said sensor pre-filtering processor filtering out a low-frequency component of the signals, and estimating a sensor bias based on the low-frequency component.

15. The system according to claim 14 wherein the sensor pre-filtering processor uses a recursive least squares technique to provide the sensor bias estimation.

16. The system according to claim 14 wherein the sensor pre-filtering processor employs a variable forgetting factor that is a value between 0 and 1 and is set based on an error between the low-frequency component of the sensor signal and an estimated bias of the sensor.

11. A system for estimating vehicle states, said system comprising:

a plurality of sensors for measuring vehicle states and providing sensor signals;

a Kalman filter observer responsive to sensor signals, said observer providing estimated vehicle states; and

a lateral velocity estimation processor responsive to some of the sensor signals and some of the estimated vehicle states, said lateral velocity estimation processor calculating a modified lateral velocity estimation signal when the vehicle is operating in a non-linear region.

12. The system according to claim 11 wherein the Kalman filter observer changes a covariance matrix in the filter when the vehicle goes from a linear to a non-linear operating region.

17. The system according to claim 11 wherein the lateral velocity estimation processor integrates the lateral velocity to modify the lateral velocity estimation signal.

18. A system for estimating vehicle states, said system comprising:

a plurality of sensors for measuring vehicle states and providing sensor signals; and

a Kalman filter observer responsive to sensor signals, said observer providing estimated vehicle states, said Kalman filter observer changing a covariance matrix in the filter when the vehicle goes from a linear to a non-linear operating region.

19. The system according to claim 18 wherein the Kalman filter observer uses a three degree-of-freedom model.

20. A system for estimating vehicle states, said system comprising:

a plurality of sensors for measuring vehicle states and providing sensor signals;

a sensor pre-filtering processor responsive to the sensor signals, said sensor pre-filtering processor filtering out a low-frequency component of the signals, and estimating a sensor bias based on the low-frequency component; and

a Kalman filter observer responsive to filtered sensor signals, said observer providing estimated vehicle states.

21. The system according to claim 20 wherein the sensor pre-filtering processor uses a recursive least squares technique to provide the sensor bias estimation.

22. The system according to claim 20 wherein the sensor pre-filtering processor employs a variable forgetting factor that is a value between 0 and 1 and is set based on an error between the low-frequency component of the sensor signal and an estimated bias of the sensor.