Path-tracking for autonomous vehicles at the limit of friction

摘要

The ability to use all of the available tire force is essential to road vehicles for emergency maneuvers and racing. As the front tires of an understeering vehicle saturate while cornering at the limit of tire-road friction, steering is lost as a control input for path-tracking. Experimental data from an autonomous vehicle show that for path-tracking at the limit of friction through steering the value of friction needs to be known to within approximately 2%. This requirement exceeds the capabilities of existing real-time friction estimation algorithms. Data collected with a professional race car driver inspire a novel control framework, with a slip angle-based control strategy of maintaining the front tires at the slip angle for which maximum tire force is attained, and longitudinal speed control for path-tracking. This approach has significantly less demanding requirements on the accuracy of friction estimation. A controller is presented to explore this concept, and experimental results demonstrate successful tracking of a circular path at the friction limit without a priori friction information.