Abstract: In order to improve the efficiency of welding robotic arm in different obstacle environments,a multi-strategy improved particle swarm algorithm for obstacle avoidance trajectory planning was proposed. The first three joints of the robotic arm were interpolated using a 6th degree polynomial function to obtain the motion trajectory. The inertia weights and learning factors of the particle swarm optimization algorithm were dynamically adjusted to balance the global and local search ability of the algorithm;the dynamic lens imaging reverse learning strategy was introduced,and the restart strategy and greedy algorithm were integrated to enhance the ability of the algorithm to jump out of the local optimum.The IRT120 robotic arm was taken as the research object and simulated by MATLAB software. The simulation results show that the improved particle swarm algorithm has significant improvements in convergence speed and optimization accuracy,and the motion trajectory is smooth without sudden changes.

Key words: robotic arm, particle swarm optimization algorithm, reverse learning strategy, obstacle avoidance trajectory planning