This paper describes a comparative study of steering and yaw moment control manoeuvres in the model predictive control and linear quadratic control approaches for path-following control of an autonomous vehicle. We present the effectiveness of the model predictive control and linear quadratic control approaches for stability control of the vehicle's lateral position and yaw angle for different control manoeuvres: two-wheel steering, four-wheel steering and direct yaw moment control. We then propose model predictive control with a feedforward controller to minimize the tracking errors of the lateral position and the yaw angle in an active front steering manoeuvre, and these are compared with the results from linear quadratic control that has a feedforward controller. Model predictive control is designed on the basis of the simple yaw-lateral motions of a single-track vehicle with a linear tyre model (i.e. a bicycle model), which is an approximation of the more realistic model of a vehicle with double-track yaw-roll motion with a non-linear tyre model (i.e. the Pacejka model). The linear quadratic controller is designed on the basis of the same approach as adopted for the model predictive controller to achieve a fair comparison. On the basis of a given trajectory, we simulate the manoeuvre of the vehicle at a low speed, a middle speed and a high speed because load transfer effects will influence the roll dynamics especially at a high speed. We also perform simulations on low-road-friction surfaces in a double-lane-change scenario with the aim of following the desired trajectory as closely as possible while maintaining the vehicle stability. The simulation results show that model predictive control through two-wheel steering and four-wheel steering with direct yaw moment control performed better in terms of trajectory tracking at a high forward speed and low road surface variation. The proposed model predictive control with a feedforward controller is shown to be effective in minimizing the trajectory tracking errors. For all control manoeuvres, model predictive control gives a better tracking performance than linear quadratic control does. In addition, when the roll dynamics are considered, model predictive control significantly improves the vehicle stability and the trajectory along the desired path.
Comparative study of autonomous path-following vehicle control via model predictive control and linear quadratic control
2015
Aufsatz (Zeitschrift)
Englisch
Path Following Control of Autonomous Underwater Vehicle Using Nonlinear Model Predictive Control
British Library Conference Proceedings | 2020
|