Design of Real-Time Obstacle Avoidance and Race Path Planning System for Unmanned Vehicle

Nowdays the local path planning has received considerable attention and achieved good effect in known environment. However, obstacles avoidance is a huge challenges in some unstructured environments and other factors. Meanwhile, the service efficiency of unmanned vehicle is greatly limited by the path planning efficiency. To solve these problems, a real-time obstacle avoidance and racing planning system for unmanned vehicles based on multi-sensor fusion is proposed which has a good balance between the success rate of driving safety and decision-making efficiency of unmanned vehicles in unknown environments. Combining the advantages of virtual force field and velocity obstacle, an obstacle avoidance model is applied to design a PID controller with subsection control for Ackerman chassis structure, which realizes efficient obstacle avoidance and high speed driving of unmanned vehicles. Finally, a dynamic weighted A* algorithm path planner with corner penalty is proposed for path planning in this work. Simulations and experimental results show that the obstacle avoidance ability of the unmanned vehicle which maintain a high speed is improved compared with the traditional tracking algorithm. And the path planning efficiency is improved compared with the traditional A * algorithm.