RRT-connect path planning algorithm based on gravitational field guidance

To address the issues of low search efficiency, rough paths, and insufficient computational efficiency in the RRT algorithm, this paper proposes a gravity-guided RRT-Connect path planning algorithm. The improved algorithm introduces an additional expansion node, enabling four random trees to perform bidirectional expansion. At the start point, goal point, and newly added intermediate expansion point, a gravitational field is applied to effectively guide the direction of node generation. This strategy significantly reduces the exploration of invalid space, thereby enhancing the algorithm's adaptability and robustness in different environments. Experimental results show that, in simple, complex, and narrow corridor simulation environments, the average path planning time of the improved algorithm is reduced by 71.61% compared to the traditional RRT-Connect algorithm, and the average number of iterations is reduced by 66.15%. The algorithm demonstrates significant advantages in path planning efficiency and search speed compared to the RRT algorithm, verifying its feasibility and effectiveness in practical applications.