Abstract: To address the issues of slow convergence speed and poor path quality of the RRT-Connect algorithm in specific scenarios of complex concave regions, an improved path planning algorithm is proposed. By introducing an optimized search range, adaptive step size, and path pruning strategy, this algorithm increases the success rate of path generation, shortens the path search time, and reduces the total path distance. Experimental results show that the improved algorithm has significant advantages in path generation efficiency and quality, with the average path planning time reduced by 11% and the average path length shortened by 30%. Both qualitative and quantitative results verify the effectiveness and superiority of the improved algorithm.