Path planning for USV based on artificial potential field approach with learning automata
Chunshui Xiong, Wei Wu, Rui Wu, Dongming Zhao · 2022
An improved artificial potential field method based on learning automata is proposed to address the problem, which method is easy to fall into local optimum while solving the path planning of unmanned vehicles. The artificial potential field method with the capability of path planning is improved by reasonably integrating reinforcement learning and artificial potential field without changing the potential field function construction. Within the virtual local stabilizing action range, by changing the angle of attraction at the target point, with the combination of optimization results of online adjustment of virtual local stabilizing action radius, angle of attraction change amount and gain parameters by learning automaton, the magnitude of repulsive force is changed to jump out of the local oscillation region, and complete the planning successfully for the case of falling into local minima. The simulation results show that the algorithm converges quickly and does not easily produce oscillation points.