Robot path planning by leveraging the graph-encoded Floyd algorithm

Abstract Path planning is an important field in the research of mobile robot technology. The so-called path planning of mobile robot means that each robot plans a path of obstacle avoidance and navigation in the same workspace to ensure that there is no collision between the robot and the robot at any time. There is no collision between robots and environmental obstacles. There are many algorithms for path planning from single robot to multi robot. However, the current robot path planning algorithms are still far from satisfactory. Thus, in this work, we propose a novel graph-based algorithm to optimize robot paths, wherein the key is a Floyd algorithm to dynamically assign weights to different paths. Firstly, we introduce the multi robot path planning methods from three aspects of multi robot cooperation, environment information and robot structure, including global path planning method, local path planning method, centralized path planning method and distributed path planning method. Next, a deep graph model by encoding the Floyd algorithm is proposed for optimizing robot paths. Finally, the development trend of multi mobile robot path planning technology is prospected and the experimental results are analyzed. Experimental results on lots of robot path datasets have shown the effectiveness and robustness of our method. Moreover, the recognized robot path sequence can guide the operation of robots in industrial area.