Method for acoustic detection of the condition of the road and the tire

• US 10,365,248 B2
• Filed: 12/15/2014
• Issued: 07/30/2019
• Est. Priority Date: 12/18/2013
• Status: Active Grant

• Current Assignee: Compagnie générale des établissements Michelin
• Original Assignee: Compagnie Generale DES Etablissements Michelin SCA
• Inventors: Paturle, Antoine, Antoni, Jerome
• Primary Examiner(s): Eyassu, Marrit
• Application Number:
  US15/103,987
• Publication Number:
  US 20160349219A1
• Time in Patent Office: 1,688 Days
• Field of Search: 73 8
• CPC Class Codes: B60C 23/00 Devices for measuring, signalling, controlling, or distributing tyre pressure or temperature, specially adapted for mounting on vehicles; Arrangement of tyre inflating devices on vehicles, e.g. of pumps or of tanks; Tyre cooling arrangements
• CPC Class Codes: B60C 23/06 Signalling devices actuated by deformation of the tyre , e.g. tyre mounted deformation sensors or indirect determination of tyre deformation based on wheel speed, wheel-centre to ground distance or inclination of wheel axle
• CPC Class Codes: B60T 2210/10 Detection or estimation of road conditions
• CPC Class Codes: G01H 3/00 Measuring characteristics of vibrations by using a detector in a fluid G01H7/00, G01H9/00, G01H11/00 take precedence
• CPC Class Codes: G01M 17/025 using infrasonic, sonic or ultrasonic vibrations
• CPC Class Codes: G01N 2291/2692 Tyres
• CPC Class Codes: G01N 29/14 using acoustic emission techniques echo of particles G01N29/046; measuring mechanical vibrations or acoustic waves in solids in general G01H1/00
• CPC Class Codes: G01N 29/4427 with stored values, e.g. threshold values
• CPC Class Codes: G01N 29/4445 Classification of defects
• CPC Class Codes: G01N 29/4472 Mathematical theories or simulation
• CPC Class Codes: G01N 29/46 by spectral analysis, e.g. Fourier analysis or wavelet analysis spectral signal processing per se G06F17/14

Abstract

In a method for determining conditions of a road and a tire, a measurement of a sound signal produced by the tire running on the road during a timeframe is recorded, and a spectral density of a power of the sound signal over a frequency interval is determined. The frequency interval is segmented into frequency bands, and each frequency band is associated with a representative data element representing a mean acoustic power measured in the frequency band. The representative data elements obtained from the measurement form variables of a vector associated with the measurement. A road condition and a tire condition are determined via a discriminant analysis of the representative data elements using a learning base. Each representative data element is obtained by finding a ratio between a mean acoustic power measured in a corresponding frequency band and a total acoustic power measured over an entirety of the frequency interval.

Claims

1. A method, implemented using a system comprising a microphone and a computer, of determining a condition of a road and a condition of a tire fitted on a vehicle travelling on the road, the method comprising steps of:

using the microphone to measure a sound signal produced by the tire running on a surface of the road during a given time frame, and recording the measurement; and

using the computer, which is connected to the microphone, to execute operations comprising;

(1) determining a spectral density of a power of the sound signal over a given frequency interval;

(2) segmenting the given frequency interval into a plurality of frequency bands of pre-determined width, each frequency band being associated with a representative data element representing a mean acoustic power measured in that frequency band, wherein the representative data elements obtained from the measurement form variables of a vector associated with the measurement; and

(3) determining a road condition and a tire condition corresponding to the vector associated with the measurement by analyzing the representative data elements using a discriminant-analysis technique and using a learning base, stored in a storage of the computer, that includes a set of vectors associated with previously recorded measurements made by the steps of the method, in known running conditions, according to modalities that each represent a given road condition or a given tire condition,

wherein each of the representative data elements forming the variables of the vector associated with the measurement is obtained by the computer by finding a ratio between (a) the mean acoustic power measured in the frequency band corresponding to the representative data element and (b) a total acoustic power measured over a whole of the given frequency interval.

2. The method according to claim 1, wherein the total acoustic power is equal to a sum of the mean acoustic powers of all the frequency bands of the given frequency interval.

3. The method according to claim 1, wherein the frequency bands are determined by dividing the given frequency interval into thirds of an octave.

4. The method according to claim 1, wherein the given time frame of the measurement is less than or equal to 0.5 seconds.

5. The method according to claim 1, wherein the given frequency interval is in a range of from 0 Hz to 20 kHz.

6. The method according to claim 1, wherein the given frequency interval is in a range of from 200 Hz to 20 kHz.

7. The method according to claim 1, wherein the modalities include a weather-condition class of modalities representing different meteorological road conditions, the weather-condition class of modalities including:

a dry condition modality, a moist condition modality, and a wet condition modality.

8. The method according to claim 1, wherein the modalities include a surface-condition class of modalities representing different conditions of a surface of the road, the surface-condition class of modalities including:

a closed condition modality (f), a medium condition modality (m), and an open condition modality (o).

9. The method according to claim 1, wherein the modalities include a wear-condition class of modalities representing different conditions of wear of the tire, the wear-condition class of modalities including:

a new condition modality (N), a half-worn condition modality (M), and a worn condition modality (U).

10. The method according to claim 1, wherein the modalities include a tread-type class of modalities representing different types of tread patterns of the tire, the tread-type modalities including:

a summer pattern modality (P), and a winter pattern modality (A).

11. The method according to claim 1, wherein the discriminant analysis technique used to analyze the representative data elements includes steps of:

determining a reduced discriminant space using the learning base, and, for each modality of a group of modalities and for each combination of a group of combinations of modalities, identifying in the reduced discriminant space an area encompassing the modality or the combination,

transforming the vector associated with the measurement to the reduced discriminant space, and, based on a location of the vector in the reduced discriminant space, associating the measurement with a probability of each modality of the group of modalities and each combination of the group of combinations of modalities, and

determining a most probable modality for each of a plurality of classes of modalities,

wherein the classes of modalities include;

a weather-condition class, a surface-condition class, a wear-condition class, and a tread-pattern class,

wherein the group of modalities include;

dry road, moist road, wet road, closed surface (f), open surface (o), medium surface (m), worn tire (U), half-worn tire (M), new tire (N), summer tread (P), winter tread (A), and

wherein the group of combinations of modalities include;

ANf, ANm, ANo, AMf, AMm, AMo, AUf, AUm, AUo, PNf, PNm, PNo, PMf, PMm, PMo, PUf, PUm, and PUo.

12. The method according to claim 11, wherein, after a determination is made that the measurement was made on a dry road, a modality belonging to the surface-condition class or the wear-condition class or the tread-pattern class is associated with the measurement.

13. The method according to claim 12, wherein:

for each modality of a class of modalities, a probability is associated with the measurement according to each combination of modalities containing the modality, and

a modality of a class of modalities having a highest probability is attributed to the measurement.

14. The method according to claim 13, wherein a diagnosis of a condition of the tire is made according to a modality in the wear-condition class of modalities or the tread-pattern class of modalities, by combining results of measurements made at different time intervals.

15. The method according to claim 1, wherein the microphone is positioned in a front part of a wheel housing located at a rear section of the vehicle.