Comparison of RRT, APF, and PSO-Based RRT-APF (PS-RRT-APF) for Collision-Free Trajectory Planning in Robotic Welding

Abstract

In robotic applications, trajectory planning is pivotal, particularly in navigating rapidly changing and dynamic environments. This paper presents a comparison of Rapidly Exploring Random Trees (RRT) and Artificial Potential Fields (APF), and Particle swarm Optimization (PSO) based RRT-APF for collision-free trajectory planning in robotic welding scenarios. The proposed methodology addresses a multifaceted set of challenges, emphasizing critical factors such as joint acceleration, obstacle avoidance, and singularity considerations. Robotic welding scenarios pose intricate challenges, demanding a multifaceted approach to ensure both precision and safety. By merging the exploration capabilities of RRT with the obstacle avoidance prowess of APF, our approach not only mitigates the drawbacks associated with local minima problems in APF but also enhances the overall efficiency of RRT. This combination enables rapid exploration and obstacle avoidance, guided by a bias force to optimize trajectory planning for enhanced robotic performance. Furthermore, the methodology extends beyond collision-free trajectories, considering key constraints to provide a holistic solution. Addressing joint, and workspace limits, proposed PS-RRT-APF approach proves collision-free trajectories while addressing specific challenges such as obstacle avoidance, joint velocities, and motion limits. Through this innovative approach, we seek to enhance the overall performance of robotic path planning, ensuring efficiency, and safety in dynamic environments.