Rapid Road Friction Estimation using Independent Left/Right Steering Torque Measurements

Publication Date



Sensing the friction coefficient between the tyres of an automobile and the road surface is a key challenge in automotive control. Accurate estimation of this parameter enables higher performance and safer operation of vehicles, whether human-operated, assisted or autonomously controlled. Previous approaches have utilised braking or lateral excitation to generate estimates but are limited by the conflict between generating sufficient excitation and preserving ride comfort and efficient operation. Further complicating the problem is the need to provide the friction estimate without adding costly sensors to the vehicle. This paper presents an approach to friction estimation that leverages separate measurements of the left and right steering torques to generate an estimate of the surface friction through direct model inversion. Uncertainty of the method is explored using experimental sensor properties and analysis of the inverted model. The method is shown to be sensitive to reduced friction coefficients and can identify sudden changes before travelling a full contact patch length. Experimental validation of the approach is demonstrated with calibrated wheel force transducers as well as low-cost sensors suitable for incorporation into production vehicles.


Mechanical Engineering