Multi-robot Multi-path Intelligent Planning Based on Universal Conflict Resolution and Free Crossing Emergence
Zeyu Zhou, Mingyang Li, Jingxi Zhang, Xiongwei Wu, Wei Tang · 2021
Multi-robot cooperative path planning is one of the most significant study areas in intelligent reconnaissance, transportation, combat and other scenarios. One of the difficulties is how to take intelligent measures to solve complex path-conflict scenarios between robots. Aiming at the above problem, this paper establishes universal conflict resolution mode and multi-path traffic rules, so that each robot can judge the characteristics of conflict scene and avoid the potential conflict scene. Firstly, an algorithm of universal conflict resolution and free crossing emergence from outside to inside (UCR-FCE) is proposed to solve the conflict problem when the number of local areas is huge, it includes extended resolution algorithms named FB, +AVRCA and BO and a mechanism of avoiding Receptor Dodger (RD) from Noumenon Dodger (ND) to the free junction; secondly, we prove that UCR-FCE can solve all the conflict scenarios with sufficient path nodes and no path deadlock, without exception; finally, a lightweight multi-robot information interaction mechanism is built. Simulation and field test of the proposed multi-robot multi-path intelligent planning algorithm were carried out. The results show that the system can effectively refer to the traffic rules, intelligently adapt to ever-changing potential conflicts, with improving success rate of path planning and multi-robot real-time communication.