Nonlinear steering controllers for the lateral dynamics stability limit

The well known nonlinear shape of the vehicle's side force characteristics is considered to design a steering controller for automatic guidance. When applying nonlinear model predictive control, which should be the best possible controller from a mathematical point of view, it turns out that a large number of controller parameters has to be defined using heuristic methods. In this paper, however, exact linearization technique is used to determine a feedback law that matches the performance of the model predictive controller with a much lower amount of computing time and only two controller parameters. The position of the vehicle fixed reference point for automatic guidance significantly affects the stability properties of the control loop, which can be proven by means of mathematical stability analysis and simulation results concerning test maneuvers at the stability limit. The capability of performing precise automatic guidance at the lateral dynamics stability limit is shown by means of a race car simulation model and a 1 : 5 scaled x-by-wire experimental prototype. During these simulations and test runs, the maximum lateral deviation from the desired trajectory was within a few percent of the vehicle width.