An integrated framework for obstacle avoidance path planning and tracking of autonomous vehicles considering risk potential fields

Abstract

Vehicle safety during driving conditions is critical, as collisions significantly contribute to traffic accidents and road congestion. This paper proposes an integrated framework combining a risk potential field and dual-layer model predictive control (MPC) for autonomous obstacle avoidance path planning and tracking. First, environmental risks are modeled through potential fields representing road boundaries, target attractions, and obstacle repulsions. Then, the upper-layer nonlinear MPC (NMPC)incorporates these potential field constraints along with vehicle dynamics to generate feasible path in real-time. Subsequently, the planned path are passed to the lower-layer MPC for accurate path tracking control. Simulation tests on the MATLAB/CarSim co-simulation platform under representative driving scenarios demonstrate that the proposed approach effectively achieves safe autonomous obstacle avoidance and stable control at high speeds,reducing collision risks and validating the method’s feasibility and effectiveness.