Acceptance–Rejection Sampling-Guided Discrete RRT for Multiagent Path Finding

Discrete rapid-exploring random tree (dRRT) methods have been proposed but remain to be improved, showing great potential for efficiently solving multiagent path finding (MAPF) problems. This article proposes a new dRRT algorithm designed to tackle complex and challenging MAPF instances that current discrete RRT approaches struggle to solve efficiently. In particular, we introduce an acceptance–rejection sampling method that efficiently guides the search for feasible paths toward destinations. Moreover, a multistep expansion strategy is proposed to improve search efficiency and path quality. Theoretical analysis regarding time complexity and probabilistic completeness is also provided. The proposed algorithm has been extensively tested on MAPF benchmark instances and verified in industrial simulations and real-world experiments, demonstrating its superior performance in handling large-scale MAPF instances and its competitiveness compared to state-of-the-art algorithms.