Method for evaluating instantaneous frequency of mechanical excitation exerted on a motor vehicle wheel

• US 7,289,893 B2
• Filed: 11/07/2002
• Issued: 10/30/2007
• Est. Priority Date: 11/28/2001
• Status: Expired due to Fees

• Current Assignee: Peugeot Citroen Automobile SA (Stellantis NV)
• Original Assignee: Peugeot Citroen Automobile SA (Stellantis NV)
• Inventors: Djama, Zahir, Gouriet, Pascal
• Primary Examiner(s): Tran; Dalena
• Application Number:
  US10/496,374
• Publication Number:
  US 20060082077A1
• Time in Patent Office: 1,818 Days
• Field of Search: 701/37, 701/38, 701/91, 73/110.8, 280/5.507, 280/5.515, 280/5.5, 280/5.503
• US Class Current: 701/37
• CPC Class Codes: B60G 17/0182 involving parameter estimation, e.g. observer, Kalman filter
• CPC Class Codes: B60G 2400/252 vertical
• CPC Class Codes: B60G 2400/91 Frequency
• CPC Class Codes: G01M 17/04 Suspension or damping

Abstract

A method for evaluating instantaneous frequency of a mechanical excitation exerted on a motor vehicle wheel by the uneven surface of a road, based on measurements of observable parameters such as those related to the instantaneous height of the chassis and/or the wheel of the vehicle. The method includes a filtering procedure producing successive estimates of the instantaneous height of the chassis and of the wheel, a lag extracting procedure, producing an estimated value of the time lag between the height estimations, and a frequency estimating procedure, producing the desired evaluation of the frequency based on the estimated time lag value.

Claims

1. A method for evaluating the instantaneous frequency of a mechanical excitation exerted by an uneven surface of a road on a wheel of a motor vehicle which travels on the road and includes a chassis connected to the wheel by suspension devices, the method including:

acquiring and elaborating at least a first measurement time series that consists of successive measurement samples representing successive values of a first observable parameter related to instantaneous height of one of the chassis and the wheel;

filtering, which uses at least the first measurement time series, to produce at least a first derived principal time series which consists of successive estimates of the instantaneous height of the chassis, and a second derived principal time series, which consists of successive estimates of the instantaneous height of the wheel,

a prior modeling procedure comprising, for a physical model associated with a system that consists of the wheel, the suspension devices, and the chassis, elaborating a first inverse transfer function which takes the first observable parameter as an input signal and produces the instantaneous height of the chassis as an output signal, and elaborating a second inverse transfer function which takes the first observable parameter as an input signal and produces the instantaneous height of the wheel as an output signal,

extracting a time lag, using at least the first derived principal series and the second derived principal series, to produce a preparatory time series that consists of successive estimated values of an instantaneous time delay between the instantaneous height of the chassis and the instantaneous height of the wheel, wherein each estimated value of the time lag is obtained by optimization of at least one function for correlation of the first derived principal time series and the second principal derived time series, and

estimating a frequency, using the preparatory time series to produce a final time series, consisting of successive estimated values of the instantaneous frequency of the mechanical excitation, wherein each frequency value is obtained, in an observation time window defined by the corresponding estimated value of the instantaneous time delay, as the frequency for which the instantaneous time delay constitutes an instantaneous phase shift.

2. The method according to claim 1, wherein at least some of the measurement samples acquired and elaborated and which pertain to an observable parameter represent relative travel of the chassis with reference to the wheel.

3. The method according to claim 1, wherein at least some of the measurement samples acquired and elaborated and which pertain to the first observable parameter represent instantaneous acceleration of the chassis.

4. The method according to claim 1, wherein at least some of the measurement samples acquired and elaborated and which pertain to the first observable parameter represent instantaneous acceleration of the wheel.

5. The method according to claim 1, wherein acquiring and elaborating comprises elaborating at least a second measurement time series, which consists of successive measurement samples that represent successive values taken by a second observable parameter that is related to the instantaneous height of one of the chassis and of the wheel.

6. The method according to claim 1, wherein the prior modeling comprises, for the physical model associated with the system that consists of the wheel, the suspension devices, and the chassis, elaborating a third inverse transfer function which takes the second observable parameter as an input signal and produces the instantaneous height of the chassis as an output signal, and elaborating a fourth inverse transfer function which takes the second observable parameter as an input signal and produces the instantaneous height of the wheel as an output signal.

7. The method according to claim 1, wherein the prior modeling comprises, for the physical model associated with the system that consists of the wheel, the suspension devices, and the chassis, elaborating a fifth inverse transfer function which takes the second observable parameter as an input signal and produces one of the instantaneous height of the chassis and of the wheel as an output signal, and elaborating a sixth inverse transfer function which takes the second observable parameter as an input signal and produces the first observable parameter as an output signal.

8. The method according to claim 1, wherein the filtering uses at least one Kalman filter.

9. The method according to claim 1, wherein the filtering uses at least two Kalman filters.

10. The method according to claim 1, wherein, in estimating a frequency, elaborating each value of the final time series as the solution, for the corresponding estimated value of the instantaneous phase shift, of at least one equation representing the physical model associated with the system which consists of the wheel, the suspension devices, and the chassis.

11. The method according to claim 1, wherein, in estimating a frequency, elaborating each value of the final time series by application of a Wigner-Ville transform or of a pseudo-Wigner-Ville transform.

12. The method according to claim 1 including controlling the suspension devices.

13. The method according to claim 1 including determining instantaneous inclination of the chassis.

14. The method according to claim 1 including determining the uneven surface of the road.

15. The method according to claim 1 including monitoring the inflation of a tire on the wheel.

16. The method according to claim 1 including monitoring mass of the vehicle.

17. The method according to claim 1 including monitoring the suspension devices of the vehicle.

18. A method for evaluating the instantaneous frequency of a mechanical excitation exerted by an uneven surface of a road on a wheel of a motor vehicle which travels on the road and includes a chassis connected to the wheel by suspension devices, the method including:

acquiring and elaborating at least a first measurement time series that consists of successive measurement samples representing successive values of a first observable parameter related to instantaneous height of one of the chassis and the wheel;

filtering, which uses at least the first measurement time series, to produce at least a first derived principal time series which consists of successive estimates of the instantaneous height of the chassis, and a second derived principal time series, which consists of successive estimates of the instantaneous height of the wheel,

extracting a time lag, using at least the first derived principal series and the second derived principal series, to produce a preparatory time series that consists of successive estimated values of an instantaneous time delay between the instantaneous height of the chassis and the instantaneous height of the wheel, wherein each estimated value of the time lag is obtained by optimization of at least one function for correlation of the first derived principal time series and the second principal derived time series, and

estimating a frequency, using the preparatory time series to produce a final time series, consisting of successive estimated values of the instantaneous frequency of the mechanical excitation, wherein each frequency value is obtained, in an observation time window defined by the corresponding estimated value of the instantaneous time delay, as the frequency for which the instantaneous time delay constitutes an instantaneous phase shift, wherein the filtering uses at least the first measurement time series to produce a first derived time series which comprises, in addition to the first derived principal time series, a first derived auxiliary time series that consists of successive variance values respectively associated with the estimates of the first derived principal time series, and a second derived time series which comprises, in addition to the second derived principal time series, a second derived auxiliary time series, that consists of successive variance values respectively associated with the estimates of the second derived principal time series, and obtains each estimated value of the time delay by optimization of at least one function for correlation of the first derived time series and the second derived time series.

19. The method according to claim 18, wherein each function for correlation of the first derived time series and the second derived time series consists of an intercorrelation function of the first derived time series and the second derived time series.

20. The method according to claim 19, including constructing each function for correlation of the first derived time series and of the second derived time series using a model based on third-order cumulants.