Multi-UGVs Collaborative Path Planning and Conflicts Eliminating in Emergent Situations

Abstract

The centralized path planning framework for multi-unmanned vehicles tends to get messy when emergency occurs, creating a series of systemic conflicts and almost impossible to return to normal spontaneously. Thus, this paper improves the Multi-UGVs Collaborative Path Planning system through emergency response planning and system recovery to eliminate the conflicts caused by emergencies. The rapid emergency response planning method is developed based on heuristic search to quickly locate a shelter with a response trajectory without disturbing other normal vehicles for each affected vehicle in real time. And then the affected vehicles can complete recovery process by introducing an asynchronous starting conflict based search (CBS) algorithm. The experiments carried out in Rviz simulation environments prove that the proposed method has good practicability and stability.