In large pipeline welding applications, manufacturing and assembly errors, along with precision control challenges during lifting and positioning, make traditional offline trajectory-based robotic welding methods ineffective. To this end, this research proposes an autonomous welding method for intersecting weld seams using a ground-rail mobile manipulator. The method integrates workpiece positioning information with the geometric model to generate a sequence of operational poses that satisfy the laser sensor’s field-of-view constraints. A comprehensive performance objective function is then constructed to address kinematic singularities, joint limits, and collision avoidance, and a multi-strategy improved particle swarm optimization (MIPSO) algorithm is employed to co-optimize the rail position and trajectory initial point. Furthermore, weld seam data obtained from the laser sensor enables further correction and optimization of the welding path through secondary trajectory planning. Experimental results show that the proposed method ensures the continuity and accuracy of the welding process. Compared to traditional experience-driven methods, it offers notable improvements in motion performance and operational efficiency, providing a systematic solution for the automatic generation of high-quality, adaptive trajectories in pipeline welding robots. • A novel and systematic solution for pipeline intersecting lines welding. • Scanning pose is automatically generated considering the field of view of the laser sensor. • An objective cost function considering both the initial point and the ground rail position. • The MIPSO algorithm is proposed to enhance the quality of searching for the optimal solution.
周静铮 et al. (Mon,) studied this question.