Collision-Free Path Planning for Dual-Arm Robots Based on Improved RRT

When using the Rapidly Exploring Random Tree (RRT) algorithm for robotic arm path planning, several challenges may arise, including slow convergence speed, low path safety, and poor smoothness. To solve these problems, this paper proposes an improved RRT algorithm. Firstly, an attractive force function is incorporated into the RRT algorithm to reduce the generation of redundant nodes, thereby decreasing the randomness of the path and improving convergence speed. Additionally, a decoupling method is proposed to prevent the occurrence of the dimension catastrophe during dual-arm path planning. Simultaneously, a layered detection method with the establishment of bounding box trees was introduced to improve collision detection efficiency and precision. Finally, the path was smoothed using Bezier curves to enhance trajectory smoothness. MATLAB simulation results demonstrate that the proposed algorithm can provide faster, safer, and smoother paths for dual-arm robots.