Path Planning Strategy of Inchworm-like Robots Based on RAG-RRT* Algorithm

In this paper, we propose a random adaptive growth RRT* (RAG-RRT*) algorithm for pipe avoidance by inchworm-like robots. Firstly, we introduce the Manhattan Distance to measure the robot joint distance in the multidimensional space and add the goal bias to improve the goal-directed exploration performance of the algorithm. Then, the node growth weights are calculated by the distance between the new node and the target point, and the joint motion trajectories are optimized by adaptively changing the growth method. At the same time, random adaptive growth is used to maintain the spatial searchability of the algorithm. Finally, the total number of nodes and path quality were compared between different algorithms by building a robot and obstacle model in the MATLAB platform, and it was verified that the RAG-RRT* algorithm effectively improved the search efficiency, reduced the number of iterations, and optimized the path quality and length.