Emergency obstacle avoidance system of driverless vehicle based on model predictive control

To improve the robustness of the emergency obstacle avoidance strategy of unmanned vehicles and reduce road traffic accidents, this paper proposes a bi-level model predictive control (MPC) for the vehicle trajectory replanning and path tracking based on obstacle avoidance function. A nonlinear MPC path planner and a linear time-varying MPC trajectory trackers are designed. The proposed approach was simulated with CarSim and MATLAB/Simulink. The simulation results show that the proposed emergency obstacle avoidance control strategy has good adaptability to low, medium and high-speed vehicles, and the vehicles can avoid obstacles stably and achieve the desired trajectory tracking, which verifies the effectiveness of the proposal.