Fast-Repair RRT* : Path Planning Algorithm for Mobile Robots in Semi-static Environment

Path planning algorithms have become particularly critical with the increasing demand for mobile robots to perform complex tasks, especially for automated navigation in scenes such as warehousing and logistics. In response to this need, this paper proposes a sampling-based path planning algorithm (Fast-Repair RRT*). The algorithm is designed to achieve path repair during local changes in the environment, and it does not require global path replanning. It is divided into two stages: initial path generation and path repair. First, an improved RRT* algorithm is used to generate the initial path. Then, a path repair method is introduced when the initial path is infeasible. This method is able to obtain feasible paths in a timely manner, and the path considers factors like path length, smoothness, and memory usage. Compared with other planning algorithms, Fast-Repair RRT* is able to generate higher quality initial paths. Through simulated experiments conducted in a semi-static environment, and the experiments confirmed the effectiveness and efficiency of the algorithm in path repair in such environments.