Automatic motion planning for complex welding problems by considering angular redundancy

Abstract Automatic motion planning in complex environment is significant in manufacturing. This paper presents an off-line collision-free motion planning algorithm by considering the task redundancy existing in manufacturing. The paper takes a typical welding technique as an example, which mainly aims at solving the complex continuous welding motion planning problems. In the proposed algorithm, the angular redundancy existing in the welding process is fully considered for planning and optimizing the welding torch path by minimizing the torch angular cost. Besides, some strategies have been introduced to improve the efficiency of the proposed algorithm, such as the heuristic region sampling strategy based on Gaussian sampling, which is adopted to guide planning. Midpoint collision checking strategy is employed to improve the efficiency of the collision checking. The proposed algorithm is very effective in solving the complex welding motion planning problems, such as in the welding environment where the weld seam is situated in the narrow passage or the dense obstacles. The experiments are carried out to verify that our proposed algorithm is feasible in the relevant scenarios. All the experimental results show that not only the proposed algorithm could find a feasible collision-free path in the different complex environments if any path exists, but also the torch angle could be optimized with the increase of iteration.