부산대학교 도서관

The ability to avoid collision with dynamic threats is critical to the safe operation of unmanned aerial vehicle (UAV). This paper presents an improved Rapidly-exploring Random Tree (RRT) based 3-D path planning algorithm for UAV to find a safe and feasible path, which the dynamic threats in a complex environment are also considered. The Artificial Potential Field (APF) algorithm is combined with the proposed algorithm to accelerate the convergence speed and optimize the found path with fewer search nodes to a shorter distance. A curve smoothing method is used for fitting the nodes to smooth the found path. Effectiveness of our proposed improved-RRT algorithm is demonstrated by numerical simulations and promising results are obtained.