Path Planning of Mobile Welding Robot Based on Fusion Algorithms of Improved RRT and TEB

To address the issues of the rapidly-exploring random tree (RRT) algorithm,such as long search times, generation of zigzag paths, and poor obstacle avoidance, as well as the drawbacks of the time-elastic band (TEB) algorithm, such as speed output jumps, a hybrid path planning algorithm for mobile welding robots that improved both RRT and TEB was proposed. A purposeful extension and key point extraction strategies were introduced to improve the RRT algorithm, and to enhance the search efficiency and smoothness of path planning. Furthermore, a minimum safety distance was introduced to improve the TEB algorithm,which enabled the mobile welding robot to effectively avoid obstacles, while increasing the constraint of target points on speed and reducing the impact on the mobile welding robot. By integrating the improved RRT and TEB algorithms, the welding robots could dynamically plan its path, and avoiding interference from dynamic obstacles. Simulations and physical validations of this approach were conducted. The results demonstrate that the path search time with the improved RRT algorithm is reduced by 58.84%, and the number of path inflection points is reduced by 73.33%. The fusion of improved RRT and TEB algorithms can increase the safe distance between the mobile welding robot and obstacles by 50%. It also provides sufficient speed buffer time before reaching the target point, which greatly enhances the stability and obstacle avoidance ability of the mobile welding robot.