Abstract: In order to make RRT^* better adapt to a variety of different complexity environments and quickly generate a smooth optimal path, this paper introduced artificial potential field method based on RRT^* algorithm, and designed RRT^* algorithm based on improved artificial potential field method. The global map was divided for partition bias sampling. Then, the node expansion method was improved, and the obstacle size factor was introduced to improve the repulsion potential field function, so the generation of new nodes was guided. Through introducing the adaptive variable step size strategy, the path points was expanded with different step sizes according to the distance from the obstacle. To make the planned path more in line with the navigation characteristics of unmanned ships, non-uniform cubic B-spline was used to smooth the path generated by the improved algorithm. In order to verify the advantages of the improved algorithm of this paper, by designing special obstacles, simple obstacles and complex obstacles, RRT, RRT^*, artificial potential field method and the algorithm of this paper was compared and analyzed.The results show that the improved algorithm generation path average length is shorter than RRT, RRT^*and artificial potential field method planning path length, the path planning efficiency is higher, and have better applicability to different obstacle environment.