Point Cloud-Based Autonomous Detection and Path Optimization of Composite Structural Weld Seams

Le résumé fourni par la source

To address the challenges of inaccurate weld seam localization and low-precision posture planning in real-world wel ding scenario, this paper proposes a composite method for weld seam recognition and trajectory optimization. A dual-constraint plane segmentation model is constructed by integrating normal vector consistency with spatial neighborhood relationships, and a dynamic weighting mechanism based on local normal vector variance is introduced to achieve adaptive and high-precision m ulti-plane segmentation. Furthermore, a three-point fitting strategy guided by principal direction projection is proposed for extr acting key weld seam points, followed by the computation of optimal intersections across multiple trajectory segments to ensur e continuity and smoothness in the welding path. Experimental results show that the proposed method enables automatic weld s eam identification, improves average trajectory accuracy by 13.5 %, and maintains welding errors within 1 mm, with a minimum error of 0.509 mm.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.