An enhanced adaptive 3D path planning algorithm for mobile robots with obstacle buffering and improved Theta* using minimum snap trajectory smoothing

Abstract

This study proposes an adaptive robot pathfinding algorithm (MS-W-Theta*) based on fused trajectory smoothing with 3D maps. Firstly, we introduce an obstacle buffer in the 3D map, which includes the height and volume of the AGV when carrying shelves. This enhancement improves the robot's path traversal in narrow spaces. Then, we propose an improved adaptive heuristic function and a cost function to optimize the total cost function of the algorithm, based on the obstacle distribution information on the 3D map. This step increases the purposefulness of the planned path. Finally, the original path is smoothed using a Minimum Snap polynomial algorithm to avoid the unstable speed or jitter phenomenon that may occur during the actual operation of the mobile robot and to improve the stability of the robot operation. The experimental results show that MS-W-Theta* exhibits a significant advantage in terms of running time, and the generated path curves better align with the requirements of real-world mobile robot scenarios. The practicality and high efficiency of the algorithm in path planning for workshop logistics robots are demonstrated through experiments, indicating a reduction in running time by 20 % and an improvement in path alignment by 15 % compared to the baseline algorithm.