Fuzzy Quadtree based Path Planner and Trajectory Smoother for A Low Cost Unmanned Aerial Vehicle

Abstract: This paper presents an effective path planning algorithm for Unmanned Aerial Vehicle (UAV) navigation based on fuzzy quadtrees. This formulation allows the user to specify the desired level of details in the planner depending on the mission complexity. A comparative study is done to investigate the advantage of the proposed fuzzy quadtree path planner over conventional quadtree path planner. It has been shown that fuzzy quadtree path planner offers a significant reduction in storage space and computation time, which are critical for low cost UAV applications. In addition to that, the optimal path predicted by the fuzzy quadtree planner was smoothened by a proposed trajectory smoother, taking vehicle's kinematic constraint into account. The user inputs the two dimensional map of obstacles to get an optimal path predicted by the fuzzy planner and then a feasible trajectory is obtained, taking UAV turn rate kinematics in consideration.