An obstacles avoidance method for serial manipulator based on reinforcement learning and Artificial Potential Field

Abstract The obstacles avoidance of manipulator is a hot issue in the field robot control. Artificial Potential Field Method (APFM) widely used path planning method, which has prominent advantages. However, APFM also some shortcomings, include inefficiency avoiding close to target or dynamic obstacles. In view shortcomings APFM, Reinforcement Learning (RL) only needs an automatic learning model continuously improve itself specified environment, makes it capable optimizing theoretically. this paper, we introduce approach hybridizing RL and solve those problems. We define concepts Distance reinforcement factors (DRF) Force (FRF) make integrated more effectively. disassemble reward function into two parts through DRF FRF, them activate different situations optimize APFM. Our method can obtain better performance finding optimal strategy by RL, effectiveness proposed algorithm verified multiple sets simulation experiments, comparative experiments physical types superior traditional other improved collisions approaching avoidance. At same time, verify practicality algorithm.