Method for detecting road and tire conditions

A method for measuring a running parameter of a vehicle traveling on a road using a discriminant analysis, the vehicle having a tire and a computing device, the method comprising the steps of:

• moving the vehicle to generate noise from the tire;
• recording a measurement of a sound signal produced by the tire traveling on a surface of the road during a time interval, the sound signal generated by the tire being measured using a microphone mounted on the vehicle;
• computing, using the computing device connected to the microphone, a spectral power density of the sound signal over a frequency interval;
• segmenting, using the computing device, the frequency interval into a plurality of frequency bands of predetermined widths and associating with each frequency band a data item representing a mean sound power of the measurement in each frequency band;
• obtaining, using the computing device, a measured vector associated with the measurement based on the data items representing the mean sound power of the measurement in each frequency band;
• obtaining, using the computing device, a condition of the road and a condition of the tire corresponding to the measured vector, using an identification method based on a learning basis, the learning basis being formed by a set of vectors associated with a plurality of measurements of sound signals produced by the tire or other tires previously recorded and performed, in accordance with the same steps as above, under known running conditions as a function of modalities, each modality representing a running condition,
• wherein the identification method identifies current road and tire condition data and comprises the following steps:
• using the learning basis to locate zones formed by each modality or one or more combinations of the modalities in a reduced discriminant space, and
• transforming the measured vector in the reduced discriminant space to a reduced measured vector to identify the current road and tire condition data;
• and
• preventing, when determining the condition of the tire, incorrect determinations resulting from modalities relatively dispersed and highly interpenetrated relative to one another by configuring the discriminant analysis to be carried out based on three combined modalities, each selected from a different class of the following three classes:
• (i) a road surface condition class, (ii) a wear class, and (iii) a tread pattern class.