Design of Control System for Self-Driving Vehicles

Abstract

Self-driving vehicles and the control system design have been undergoing rapid changes in the last decade and affecting the concept and behaviour of human traffic. However, the control system design for autonomous driving vehicles is still a great challenge since the real vehicles are subject to enormous dynamic constraints depending on the vehicle physical limitations, environmental constraints and surrounding obstacles. This paper presents a new scheme of nonlinear model predictive control subject to softened constraints for autonomous driving vehicles. When some vehicle dynamic limitations can be converted to softened constraints, the model predictive control optimizer can be easier to find out the optimal control action. This helps to improve the system stability and the application for further intelligent control in the future. Simulation results show that the new controller can drive the vehicle tracking well on different trajectories amid dynamic constraints on states, outputs and inputs.