Multi-robot path planning and cooperative obstacle avoidance based on Monte-Carlo tree search

With the rapid development of artificial intelligence technologies, multi-robot cooperative processing tasks are becoming more and more common in various applications. However, traditional single-robot planning algorithms show obvious limitations in dealing with complex social interactions and dynamic environments in the real world. To address this challenge, this paper applies Monte Carlo Tree Search (MCTS) to multi-robot path planning and proposes a collaborative obstacle avoidance motion planning algorithm based on MCTS. Specifically, the algorithm prunes the Monte Carlo tree by excluding cases that do not meet the robot speed constraints and potential collisions, which significantly improves the computational speed and success rate of the algorithm, making it applicable to the multi-robot path conflict problem. In addition, we design a policy space that conforms to the robot's operation and a cost function that meets the requirements of robot path planning to plan an efficient cooperative path for each robot. Through a series of simulation experiments, we demonstrate the algorithm's path planning and obstacle avoidance performance in a variety of conflict scenarios. The results show that the method is highly cooperative and effective in complex environments.