Simultaneous Stabilization and Tracking of Basic Automobile Drifting Trajectories
Professional drivers in `drifting' competitions are able to precisely negotiate a specified course at high sideslip angles while operating in an unstable region of state-space. Studying this practice could provide insight into autonomous car control during emergency maneuvers that excurse outside stable handling limits. This paper presents a simple and physically insightful controller for autonomous drifting with simultaneous tracking of a reference path. A feasible reference trajectory is treated as a sequence of unstable drifting equilibrium points, and a basic example is generated from vehicle parameters using a four-wheel model with steady-state weight transfer. Lookahead error and sideslip are chosen as reference states, and a controller for tracking both objectives around an equilibrium point is derived using a simpler single-track model. Experiments on the rear-wheel drive MARTY test vehicle demonstrate good tracking performance of both objectives even at values of sideslip as high as 45 degrees.
2016 IEEE Intelligent Vehicles Symposium