Obstacle Avoidance of Autonomous Vehicles with CQP-based Model Predictive Control

In this paper, an approach for real time obstacle avoidance of autonomous vehicles is presented. A model predictive control (MPC) scheme based on convex quadratic programming (CQP) is developed to generate safety trajectories. To reduce the computational burden in optimizing the performance index of MPC, linear time-varying MPC is adopted and a unique single dimension artificial potential fields (SDAPF) method to utilize the obstacle information is proposed. Autonomous vehicles with proposed method can track the desired path if there is no obstacle on it and avoid both static and dynamic obstacles if the path is occupied. Simulation results show the validity of the approach and its superior real time performance, which is critical to autonomous vehicles.