Model learning of the tire-road friction slip dependency under standard driving conditions

The maximum tire-road friction coefficient is a quantity allowing to increase significantly the performance of the driver-assistance systems. Unfortunately, this value is particularly difficult to estimate under standard driving conditions. Therefore, this paper focuses on a parametric method based on Monte-Carlo Markov Chain techniques to determine it accurately. Our solution combines the advantages of both a maximum likelihood method and an adaptive Metropolis algorithm to determine the tire-road friction coefficient. This approach requires friction measurements and a tire model representing the friction as a function of the slip ratio. After describing the theoretical aspects of the method, this paper focuses on its practical implementation. Several results obtained from both simulated data and real data are given to demonstrate the efficiency of the approach.

Automated summary

This paper presents a parametric method based on Monte-Carlo Markov Chain techniques to accurately determine the maximum tire-road friction coefficient. The method combines the advantages of a maximum likelihood method and an adaptive Metropolis algorithm, requiring friction measurements and a tire model for calculation.