Path Planning for Coal Mining Masonry Robots Combined With Trajectory Optimization

Abstract

In underground coal mining, the efficiency of masonry robots is hindered by complex environmental conditions and pose constraints. This study proposes a novel path planning algorithm combining an improved Rapidly-exploring Random Tree (RRT) with Particle Swarm Optimization (PSO), followed by trajectory optimization under mechanical constraints to identify the time-optimal path. The improved RRT incorporates dynamic sampling regions and tree reorganization to reduce redundancy and enhance efficiency. A dynamic step length strategy is also introduced to address obstacle avoidance in complex underground environments, ensuring robotic arm safety. The modified PSO algorithm is then used for path planning and trajectory optimization, incorporating obstacle avoidance and pose constraints. Simulation results show that the integrated algorithm significantly reduces path length, sampling points, and search time compared to traditional RRT, RRT*, and informed RRT*. Additionally, trajectory optimization with PSO, considering joint posture constraints, reduces operation time by approximately 13% compared to ant colony optimization. This research provides key technical insights for improving the efficiency and safety of masonry robots in coal mining.