Periodic Multi-Agent Path Planning

Multi-agent path planning (MAPP) is the problem of collision-free trajectories from start to goal locations for a team agents. This work explores relatively unexplored setting MAPP where streams agents have go through starts and goals with high throughput. We tackle this by formulating new variant called periodic in which timing agent appearances periodic. The objective find plan, set that can use repeatedly over periods, periods are as small possible. To meet objective, we propose solution method based on constraint relaxation optimization. show plans once found be used more practical case stream appear at random times. confirm effectiveness our compared baseline methods terms throughput several scenarios abstract autonomous intersection management tasks.