In order to solve the problems of low success rate and low efficiency of 6R manipulator path planning in a complex environment,an improved APF-RRT fusion algorithm was proposed.For the unreachable problem of the artificial potential field (APF),the repulsive adjustment factor was introduced into the repulsive function,so that the repulsive force of obstacles on the manipulator gradually decreased when the robot arm was near the target point,then the robot arm could reach the target point smoothly.Aiming at the problem of rapidly exploring random trees (RRT) algorithm with strong randomness,a target oriented strategy was proposed to make the manipulator extended to the target point with a certain probability.When the artificial potential field method fell into local optimal,the improved RRT algorithm was used for path planning.When the local minimum was skipped,the artificial potential field method was switched to continue path planning.Finally,the simulation results show that the improved APF-RRT algorithm can adapt to various complex environments,and has the advantages of short planning time and high planning success rate compared with traditional APF and RRT algorithms,the problem of unreachable target of APF method and local minimum is solved effectively.Finally,the feasibility of the improved APF-RRT fusion algorithm is verified by JAKA robot experiment platform.