Research on the AGV Global Path Planning Algorithm Based on FI-RRT*

To address the issue of slow speed for AGV during global planning tasks, we propose the FI-RRT* algorithm based on the Informed-RRT* algorithm. In the sampling phase of the algorithm, we adjust the distribution probability of randomly sampled points using a Gaussian distribution function, which saves memory and increases node utilization. In the node reselection phase, we adopt an improved elliptical convergence domain strategy to further reduce the size of the convergence domain, thereby enhancing the efficiency of the reselection step. Experimental results obtained through dual-platform experiments using python and ROS demonstrate that the FI-RRT* algorithm has a higher planning speed.