<p>Autonomous vehicle collision avoidance technology is crucial for improving road safety and transportation efficiency. In this paper, a novel integrated vehicle collision avoidance structure is proposed, which includes a path planning layer and a path tracking control layer. In the planning layer, an improved double exponential three-dimensional virtual risk potential field (RPF) is established, taking into account the real-time planning and path smoothness. In the control layer, the fuzzy model predictive controller (F-MPC) algorithm with a soft constraint on the tire slip angle is established, taking into account the steadiness and tracking accuracy of the vehicle at high speed. The introduction of the soft constraint of tire slip angle ensures the steadiness of the vehicle and the comfort of passengers at high speed. The introduction of the fuzzy algorithm not only improves the adaptability of different speeds of the vehicle to the time–domain parameters, but also further improves the tracking accuracy and guarantees the driving safety. In order to verify its practical feasibility, this paper carries out a joint simulation based on Matlab/Simulink/Carsim, and the results show the effectiveness of the proposed method.</p>