Abstract:In this paper, we address the problem of both priority and path planning for multiple robots. Many relevant studies showed that priority assigned to robots has a significant impact on the planning results, such as whether a solution can be found or the global cost. We propose a double-layer structure that combines priority planning and path planning. A modified A* algorithm is introduced to get collision-free paths as a path planner layer. A tree-like searching method that conducting in configuration space of priority is designed as a priority planner layer, while the mode communicating with the path planner layer is also defined to abstract heuristic information to guide searching direction. This new approach is implemented in simulation and represents good performance both in success rate and global paths cost.

Combined Priority and Path Planning Through a Double-layer Structure for Multiple Robots.