Path Planning of Load Transportation by Multi-Quadrotor Formation Based on RRT-APF Method

Abstract

In this work, we propose an integration algorithm of rapidly-exploring random trees method and artificial potential field method in path planning for a multi-quadrotor unmanned aerial vehicles formation with a load. Specifically, we establish potential field to guide the stretch direction of branches and extract the key points of the final planned path for optimization. For the lack of intelligence of the artificial potential field method, the sampling method gets improvements both on that and rapidity. In the same way, the artificial potential field method also makes up for the lack of planned path's smoothness with the sampling method. Aiming at the complicated structure of the transporting system, we set strict safety constraints to ensure the success of obstacle avoidance. In addition, an algorithm of searching along the wall is proposed to fit in the scene with unknown information obstacles. The methods' feasibility and effectiveness are verified by simulation results and relevant comparative analyses.