面向砌筑机器人机械臂路径规划的改进Q-RRT*算法OA

Aiming at the limitations of the traditional Q-RRT* algorithm in path quality and convergence speed,an improved Q-RRT* algorithm was developed.A dual-random-tree target-biased sampling strategy was introduced to guide sampling points toward the target region.A k-dimensional tree was adopted to construct spatial indexes and improve the efficiency of nearest neighbor search.The bidirectional midpoint optimization strategy and the second-order Bézier curve fitting method was integrated to improve the path smoothness and feasibility.Multiple optimization mechanisms work synergistically to improve both the path planning efficiency and quality.To validate the algorithm performance,comparative tests were conducted in the 2D and 3D simulation environments on the MATLAB platform,and further deployment and verification were implemented on a digital twin platform for masonry robots and a physical UR10 manipulator.The results showed that the improved Q-RRT* algorithm effectively suppressed redundant exploration in invalid regions,reduced the time complexity of nearest neighbor search from O(n)to O(log n),and enhanced the path continuity.Compared with the traditional Q-RRT* algorithm,in 2D and 3D simulation environments,the improved algorithm reduced the initial path computation time by 88.3%and 90.8%,decreased the number of iterations by 86.8%and 84.8%,and shortened the path length by 4.7%and 6.3%,respectively.The proposed algorithm demonstrates superior robustness,good engineering applicability,and broad application potential.