Decentralized motion planning for multi quadrotor with obstacle and collision avoidance

In this paper, a decentralized motion planning approach for multi-quadrotor autonomous navigation is developed in environments populated with obstacles. We propose a priority-based RRT algorithm, which generates an online global path for each quadrotor with obstacle avoidance. Furthermore, the front-end paths uncross by assigning priorities and sharing global states in the swarm. The security and clearance of the trajectory are improved by B-spline optimization, where inter-collision-free is achieved by formulating the collision risk as a penalty term of the cost function. The efficiency of the developed algorithm is verified through numerical simulation.