Dynamic Node Allocation-Based Multirobot Path Planning

Abstract

The previous multirobot path planning algorithm stipulated that more than two channels must be included in the map, causing the algorithm to malfunction in more demanding environments such as narrow passages or crossroads. In contrast, a path planning method based on heuristic search is proposed in this paper. The method truncates the path of robots and disables inverse edges so that it enables multirobots to run in a coordinated way under space-limited circumstances. First, real-time communication between the server and the robots is established. Then, the server uses heuristic search to plot a short path for the robot according to the obtained robot position and state information, during which the inverse edges of the node sequence allocated to other robots and the nodes occupied by stationary robots are not allowed to be visited. Finally, since the server has searched the paths for all the robots, the duplicated nodes are eliminated while the remnant nodes are sent to the corresponding robots. The above process loops until each robot reaches the target position. Experimental results show that this method improves the traffic efficiency of multirobots in narrow passages and allows robots to operate in an orderly manner.