Randomized multi-goal path planning for Dubins vehicles

In multi-goal path planning, the task is to find a sequence to visit a set of target locations in an environment using the shortest path. Finding the order can be achieved by solving an instance of the Traveling Salesman Problem (TSP). This requires to determine paths and their lengths between the individual targets, which is solved using robotic path planning. In this paper, we propose a randomized planner for multi-goal path planning for non-holonomic mobile robots. Multiple trees are constructed simultaneously from the targets and expanded by collision-free configurations until they touch each other or obstacles. The trees are expanded using Dubins maneuvers. Therefore, the resulting trajectories satisfy the kinematic constraints of the mobile robots and they can also be used by Unmanned Aerial Vehicles flying at a constant altitude. The efficiency of the proposed planning approach is demonstrated in the multi-goal path planning for environments with tens of targets and compared to state-of-theart approaches.