Vehicle slip angle measuring method and a device therefor

• US 5,579,245 A
• Filed: 02/04/1994
• Issued: 11/26/1996
• Est. Priority Date: 03/17/1993
• Status: Expired due to Fees

• Current Assignee: Mitsubishi Jidosha Kogyo Kabushiki Kaisha
• Original Assignee: Mitsubishi Jidosha Kogyo Kabushiki Kaisha
• Inventors: Kato, Manabu
• Primary Examiner(s): Trammell, James P.
• Assistant Examiner(s): Wachsman, Hal D.
• Application Number:
  US08/191,082
• Time in Patent Office: 1,026 Days
• Field of Search: 364/559, 364/424.01, 364/424.03, 364/424.04, 364/424.05, 364/425, 364/426.04, 364/426.01, 364/426.03, 180/140, 180/142, 180/141, 180/79.1, 180/79, 180/197, 395/903, 395/905, 395/913, 395/2.41
• US Class Current: 702/150
• CPC Class Codes: B60T 2230/02 Side slip angle, attitude angle, floating angle, drift angle
• CPC Class Codes: B62D 6/00 Arrangements for automatically controlling steering depending on driving conditions sensed and responded to, e.g. control circuits means for initiating a change in direction B62D1/00; steering valves B62D5/06; combined with means for inclining the vehicle body or wheels on bends B62D9/00
• CPC Class Codes: B62D 7/159 characterised by computing methods or stabilisation processes or systems, e.g. responding to yaw rate, lateral wind, load, road condition

Abstract

A vehicle slip angle measuring device includes an approximate calculation block for deriving an approximate value of a vehicle centroid slip angle by use of an approximate expression derived from a linear two-degree-of-freedom vehicle model. The approximate value is derived based on outputs from a steering wheel sensor, vehicle velocity sensor and yaw angular velocity sensor. The device further includes a preprocessing block for preprocessing outputs from the above three sensors, longitudinal acceleration sensor and lateral acceleration sensor to create input information. Finally, the device includes a learned neural network for receiving the approximate value and input information and for outputting a correction value corresponding to a deviation between an actual slip angle and the approximate value. The correction value from the neural network is added to the approximate value from the approximate calculation block to derive a precise vehicle centroid slip angle.

Claims

1. A vehicle slip angle measuring method, comprising steps of:

(a) detecting, using at least one sensor of the vehicle, a steering state of front wheels of a vehicle;

(b) detecting, using at least one sensor of the vehicle, a motion state of the vehicle;

(c) deriving an approximate value of a vehicle slip angle at a centroid of the vehicle based on the front wheel steering state and the motion state, respectively detected in said steps (a) and (b);

(d) supplying the front wheel steering state and the motion state, respectively detected in said steps (a) and (b) and the approximate value derived in said step (c) to a learned neural network, to estimate a correction value corresponding to a deviation between the approximate value and an actual value of the vehicle slip angle at the centroid of the vehicle; and

(e) controlling the vehicle based upon a value derived by correcting the approximate value derived in said step (c), by using the correction value estimated in said step (d), as the vehicle slip angle at the centroid of the vehicle body.

2. A vehicle slip angle measuring method according to claim 1, further including a step of:

(f) effecting a learning operation for a neural network to construct the learned neural network, said step (f) including the sub-steps of;

(f1) detecting a steering state of the front wheels of the vehicle;

(f2) detecting a motion state of the vehicle;

(f3) deriving an approximate value of a vehicle slip angle based on the front wheel steering state and the motion state, respectively detected in said sub-steps (f1) and (f2);

(f4) supplying the front wheel steering state and the motion state, respectively detected in said sub-steps (f1) and (f2) and the approximate value derived in said sub-step (f3) to the neural network, to estimate a correction value corresponding to a deviation between the approximate value and an actual value of the vehicle slip angle;

(f5) detecting an actual value of the vehicle slip angle;

(f6) deriving a deviation between the actual value of the vehicle slip angle detected in said sub-step (f5) and the approximate value derived in said sub-step (f3); and

(f7) supplying the front wheel steering state and the motion state, respectively detected in said sub-steps (f1) and (f2), the approximate value derived in said sub-step (f3), and a difference between the correction value estimated in said sub-step (f4) and the deviation calculated in said sub-step (f6) to the neural network, and changing an input/output characteristic of the neural network such that the difference is reduced to a minimum.

3. A vehicle slip angle measuring method according to claim 1, wherein said step (a) includes detecting a steering angle of the front wheels of the vehicle;

wherein said step (b) includes detecting a vehicle velocity and yaw angular velocity of the vehicle; and

wherein said step (c) includes deriving the approximate value based on the front wheel steering angle, the vehicle velocity and the yaw angular velocity detected in said steps (a) and (b) by using an approximate expression which expresses an approximate value of the vehicle slip angle as a function of vehicle velocity, front wheel steering angle and yaw angular velocity of the vehicle, said approximate expression being derived from an equation of motion of the vehicle which is in turn derived from a linear mathematical vehicle model.

4. A vehicle slip angle measuring method according to claim 1, further including the step of:

(g) deriving at least one quantity of state from the front wheel steering state detected in said step (a) and/or the motion state detected in said step (b);

the vehicle slip angle being expressed by said at least one quantity of state or a sum thereof;

each of said at least one quantity of state being expressed as a function of a corresponding one or ones among the front wheel steering state detected in said step (a) and the motion state detected in said step (b);

each of said at least one quantity of state being input, in said step (d), to the learned neural network as the detected front wheel steering state and/or the detected motion state.

5. A vehicle slip angle measuring method according to claim 4, further including the step of:

(h) storing detected front wheel steering states, sequentially detected in said step (a) effected at a plurality of present and/or past timings, as time-series data, the time-series data being input, in said step (d), to the learned neural network as the detected front wheel steering state.

6. A vehicle slip angle measuring method according to claim 5, wherein said step (a) includes detecting a steering wheel angle of the vehicle;

wherein said step (g) includes deriving front wheel steering angles at said plurality of timings as said at least one quantity of state, based on steering wheel angles sequentially detected in said step (a); and

wherein said step (h) includes storing front wheel steering angles at said plurality of timings as said time-series data.

7. A vehicle slip angle measuring method according to claim 5, wherein said step (a) includes detecting a steering wheel angle of the vehicle;

wherein said step (b) includes detecting a yaw angular velocity, longitudinal acceleration and lateral acceleration of the vehicle and a vehicle velocity;

wherein said step (g) includes deriving at least one of the following;

(lateral acceleration-vehicle velocity×

yaw angular velocity), (vehicle velocity ×

yaw angular velocity), (yaw angular velocity÷

vehicle velocity), (yaw angular acceleration), (longitudinal acceleration×

yaw angular velocity) and (front wheel steering angles at said plurality of timings), as said at least one quantity of state, based on the yaw angular velocity, longitudinal acceleration, lateral acceleration and vehicle velocity detected in said step (b) and the steering wheel angle detected in said step (a); and

wherein said step (h) includes storing front wheel steering angles at said plurality of timings as said time-series data.

8. A vehicle slip angle measuring method according to claim 4, wherein said step (b) includes detecting at least one of a yaw angular velocity, longitudinal acceleration and lateral acceleration of the vehicle and a vehicle velocity;

and

wherein said step (g) includes deriving said at least one quantity of state each expressed as a function of a corresponding one or ones of yaw angular velocity, longitudinal acceleration, lateral acceleration and vehicle velocity, based on at least one of the yaw angular velocity, longitudinal acceleration, lateral acceleration and vehicle velocity detected in said step (b).

9. A vehicle slip angle measuring method according to claim 8, wherein said at least one quantity of state includes at least one of the following:

(lateral acceleration-vehicle velocity ×

yaw angular velocity), (vehicle velocity×

yaw angular velocity), (yaw angular velocity÷

vehicle velocity), (yaw angular acceleration), and (longitudinal acceleration×

yaw angular velocity).

10. A vehicle slip angle measuring method according to claim 1, wherein said step (e) includes outputting a value obtained by adding together the approximate value derived in said step (c) and the correction value estimated in said step (d) as said vehicle slip angle.

11. A vehicle slip angle measuring device comprising:

first detection means for detecting a steering state of front wheels of a vehicle;

second detection means for detecting a motion state of the vehicle;

calculation means for deriving an approximate value of a vehicle slip angle at a centroid of a body of the vehicle based on the front wheel steering state and the motion state respectively detected by said first and second detection means;

a learned neural network for estimating a correction value, corresponding to a deviation between the approximate value and an actual value of the vehicle slip angle at the centroid of the vehicle body, based on the front wheel steering state and the motion state respectively detected by said first and second detection means and the approximate value derived by said calculation means; and

correction means for outputting a value derived by correcting the approximate value derived by said calculation means, by using the correction value estimated by said learned neural network, as the vehicle slip angle at the centroid of the vehicle body.

12. A vehicle slip angle measuring device according to claim 11, wherein said learned neural network is constructed by supplying a neural network with the front wheel steering state and the motion state respectively detected by said first and second detection means, the approximate value derived by said calculating means, and a difference between the correction value and a deviation of the approximate value from an actual value of the vehicle slip angle, and by changing an input/output characteristic of the neural network such that the difference is reduced to a minimum.

13. A vehicle slip angle measuring device according to claim 11, wherein said first detection means includes steering angle detection means for detecting a steering angle of the front wheels of the vehicle;

wherein said second detection means includes vehicle velocity detection means for detecting a vehicle velocity, and yaw angular velocity detection means for detecting a yaw angular velocity; and

wherein said calculating means is so constructed as to derive said approximate value based on the front wheel steering angle, the vehicle velocity and the yaw angular velocity respectively detected by said steering angle detection means, said vehicle velocity detection means and said yaw angular velocity detection means, by using an approximate expression expressing an approximate value of the vehicle slip angle as a function of front wheel steering angle and yaw angular velocity of the vehicle and vehicle velocity, said approximate expression being derived from an equation of motion of the vehicle which is in turn derived from a linear mathematical vehicle model.

14. A vehicle slip angle measuring device according to claim 11, further including:

processing means for deriving at least one quantity of state from the front wheel steering state detected by said first detection means and/or the motion state detected by said second detection means, and for outputting the at least one quantity of state to said learned neural network, the vehicle slip angle being expressed by said at least one quantity of state or a sum thereof, each of said at least one quantity of state being expressed as a function of a corresponding one or ones of the front wheel steering state detected by said first detection means and the motion state detected by said second detection means.

15. A vehicle slip angle measuring device according to claim 14, wherein said processing means includes memory means for storing detected front wheel steering states, sequentially detected by said first detection means at a plurality of present and/or past timings, as time-series data, said processing means being operable to output the time series data stored in said memory means to said learned neural network as the detected front wheel steering state.

16. A vehicle slip angle measuring device according to claim 15, wherein said first detection means includes a steering wheel angle sensor for detecting a steering wheel angle of the vehicle;

wherein said calculating means derives front wheel steering angles at said plurality of present and or past timings, as said at least one quantity of state, based on steering wheel angles sequentially detected by said steering wheel angle sensor; and

said memory means stores the front wheel steering angles at said plurality of timings as said time-series data.

17. A vehicle slip angle measuring device according to claim 15, wherein said first detection means includes a steering wheel angle sensor for detecting a steering wheel angle of the vehicle;

wherein said second detection means includes a yaw angular velocity sensor for detecting a yaw angular velocity of the vehicle, a longitudinal acceleration sensor for detecting a longitudinal acceleration of the vehicle, a lateral acceleration sensor for detecting a lateral acceleration of the vehicle and a vehicle velocity sensor for detecting a vehicle velocity;

wherein said calculating means derives at least one of the following;

(lateral acceleration-vehicle velocity×

yaw angular velocity), (vehicle velocity ×

yaw angular velocity), (yaw angular velocity÷

vehicle velocity), (yaw angular acceleration), (longitudinal acceleration×

yaw angular velocity) and front wheel steering angles at said plurality of present and/or past timings, as said at least one quantity of state, based on the yaw angular velocity, the longitudinal acceleration, the lateral acceleration, the vehicle velocity and the steering wheel angle respectively detected by said yaw angular velocity sensor, said longitudinal acceleration sensor, said lateral acceleration sensor, said vehicle velocity sensor and said steering wheel angle sensor; and

wherein said memory means stores the front wheel steering angles at said plurality of timings as said time-series data.

18. A vehicle slip angle measuring device according to claim 14, wherein said second detection means includes at least one of a yaw angular velocity sensor for detecting a yaw angular velocity of the vehicle, a longitudinal acceleration sensor for detecting a longitudinal acceleration of the vehicle, a lateral acceleration sensor for detecting a lateral acceleration of the vehicle, and a vehicle velocity sensor for detecting a vehicle velocity, and

wherein said processing means includes calculation means for deriving said at least one quantity of state, each expressed as a function of a corresponding one or ones of yaw angular velocity, longitudinal acceleration, lateral acceleration and vehicle velocity, based on at least one of the yaw angular velocity, longitudinal acceleration, lateral acceleration and vehicle velocity detected by at least one of said sensors.

19. A vehicle slip angle measuring device according to claim 18, wherein said calculation means derives at least one of the following:

(lateral acceleration-vehicle velocity×

yaw angular velocity), (vehicle velocity×

yaw angular velocity), (yaw angular velocity+vehicle velocity), (yaw angular acceleration), and (longitudinal acceleration ×

yaw angular velocity) as said at least one quantity of state.

20. A vehicle slip angle measuring device according to claim 11, wherein said correction means includes adding means for adding together the approximate value derived by said calculating means and the correction value estimated by said learned neural network, said adding means being operable to output a result of addition as said vehicle slip angle.