Unmanned Vehicle Path Planning Based On Dynamic Potential Field Method

Path planning algorithm for unmanned vehicle in dynamic environment becomes an urgent problem to be solved, with the development of intelligent transportation, the transformation of the automotive industry and the demand for military and other aspects.After analyzing and optimizing classical artificial potential field method, unmanned vehicle path planning based on dynamic potential field method was proposed. After solving problems of unable to reach their destination and local minimun trap, path planning on dynamic potential field method was proved with softwares and experiments were done with IN-RE robot designed by shanghai partner and future robot company.Classic artificial potential field method and the main problems of this method were analysised deeply. Detailed simulation was done after solving the two problems with improving potential field modeling of repulsion and virtual water method. Simulation results show that the problems of classic artificial potential field method have been solved successfully with the improved algorithm.In dynamic environment, considering the relative position and relative velocity between target point and unmanned vehicle, gravitational potential field and gravitational function were established. Considering the relative position, relative velocity and relative acceleration factor between unmanned vehicle and obstacles, repulsive potential field and repulsion function were established. Based on the “divide and rule” strategy, the running environment of unmanned vehicle was decomposed into different scenarios. According to the environmental information obtained by sensors, current scene of unmanned vehicle was judged and corresponding action was selected according to current scene and predefined scenario- motion rule. Simulations of static and dynamic environment were done. Results show that unmanned vehicles can reach the target point safely and the improved artificial potential field method could be applied to dynamic environment.The kinematics model of IN-RE wheeled robot was established and virtual feedback with hyperbolic characteristics was introduced. The trajectory tracking algorithm of unmanned vehicle was studied based on Back-stepping and asymptotic stability of control system is proved by Lyapunov theory. Simulation of tracking circular trajectory was done and simulation results show that the algorithm is capable of quick and precise tracking and global stability. Experiments were done in the environment of dynamic and obstacles near the target point. The results show that IN-RE robot could avoid the dynamic obstacles and reach the target point successfully. What’s more, results also verify that the improved artificial potential field method is applied to the dynamic environment.