Multi-Robot Trajectory Planning with Hybrid Maps Based on the MADER Framework

In complex outdoor scenarios, ground robots require safe and efficient autonomous navigation, which relies on fast and accurate trajectory planning. However, current methods often fail to meet the high complexity of the environment and the high efficiency of operation simultaneously. To address this issue, we propose a multi-ground robot trajectory planning method based on the MADER framework combined with a hybrid map. By creating hybrid maps to determine passability, and representing obstacles/robots with minimum volume polyhedra, the method uses the separating planes between polyhedra and the passability constraints of the point cloud map as decision variables for trajectory planning and optimization. It effectively avoids collisions with dynamic obstacles or other robots in rugged terrain. The simulation results indicate that our method is safe and efficient, significantly reducing the robots' running and stopping times, as well as the travel distance.