Aiming at randomness problem that the path planning sampling of the two-way fast-expanding random tree（ RRT-Connect） has in the complex static environment

an improved algorithm combining artificial potential field method and RRT-Connect algorithm（ PRRT-Connect） is proposed to optimize sampling and scaling strategies.The selective heuristic factor is used to search the tree row expansion of the RRT-Connect algorithm. The planned path is trimmed according to the performance constraints of the UAV and the final path is performed using the third-order Bezier curve

so that the path planned by the improved algorithm is more close to the optimal path. The research results show that the improved algorithm has the characteristics of high efficiency

good convergence performance and strong planning ability

and it can avoid unexpected obstacles

which can well solve the path planning problem of UAVs in complex environments.