Autonomous Vehicle Path Following and Speed Control Based on Model Predictive Control

The current research on path following mainly focuses on lateral control under the assumption of uniform speed, ignoring the influence of longitudinal motion. In this paper, we proposed a combined lateral and longitudinal control structure based on model predictive control (MPC) theory to achieve path following while controlling the speed. For the lateral MPC path following, we designed the objective function and dynamics constraints, and dynamically adjusted the weight matrix according to the curvature of reference path. The simulation results illustrate that the lateral following error is significantly reduced. The longitudinal controller based on MPC uses the desired speed to calculate the desired acceleration, and the safe reference speed is generated with the speed decision model. Finally, taking the longitudinal speed as the coupling point, the lateral and longitudinal controllers are combined. The co-simulation is made to demonstrate the accurate following performance of the proposed MPC controllers and the stability of vehicle motion.