Dynamic tire-pavement friction prediction with an integrated sensing-modeling approach

Abstract

This study proposed an intelligent tire solution to predict tire-pavement friction from tire sensors using an integrated modeling-sensing approach. A laboratory platform is built to conduct dynamic tire tests under different operating parameters and surface conditions. Pressure-based sensors were embedded in the tire tread rubber to measure local forces on the tire contact patch. Physics-based models are built to interpret the friction generation mechanisms and predict the global friction force from sensor measurements. The tire−pavement interaction model consists of a Brush model for tire−pavement contact, a flexible ring model for tire stress and strain, and energy dissipation theory. The flexible ring model parameters are first calibrated with tire load−deflection curves. The feasible dynamic friction coefficients and the deformed tire profile were then solved using an interactive process among the three models using sensor measurements. Finally, the predicted friction forces were compared with the reference measurements from load cells to evaluate the prediction accuracy. The results confirmed the capability of smart tire sensing for estimating tire−pavement friction coefficients at various slip ratios under different surface conditions, which shows the potential for friction-informed vehicle control and safe driving.