Scholarlink

<p>To address the challenges of irregular geometry, significant curvature variations, and disordered normal vector distribution in aviation Invar steel S-type molds, this paper proposes a welding trajectory generation method based on multi-level trajectory fitting and adaptive connection. To address issues such as high trajectory fitting error rates and trajectory gaps caused by irregular geometric shapes and significant curvature variations, the model point cloud is first segmented into regions. Subsequently, trajectory points are extracted using slicing operations and domain searches, and trajectory fitting is performed via Euclidean clustering. After obtaining a simple trajectory, an adaptive connection mechanism is introduced to enhance the algorithm’s practicality, thereby translating the algorithm’s intended outcomes into actual results. The proposed algorithm achieves a fitting accuracy exceeding 90%, with smoothness and average <i>Z</i>-direction values below 0.1, objectively demonstrating the high accuracy and stability of the trajectory fitting method presented herein. This study provides a feasible solution for automated welding of aviation Invar steel molds and offers new insights for the development of robotic welding trajectory planning.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on welding trajectory planning for aviation Invar steel S-type mold based on multi-level trajectory fitting and adaptive connection

  • 期刊论文

  • 发表日期：

    2025年11月26日

查看全文
  • Dongling Yu,
  • Xianqi Liao,
  • Chenggui Liao,
  • Sheng Liao,
  • Zengguang Lai,
  • Chao Bao

摘要

To address the challenges of irregular geometry, significant curvature variations, and disordered normal vector distribution in aviation Invar steel S-type molds, this paper proposes a welding trajectory generation method based on multi-level trajectory fitting and adaptive connection. To address issues such as high trajectory fitting error rates and trajectory gaps caused by irregular geometric shapes and significant curvature variations, the model point cloud is first segmented into regions. Subsequently, trajectory points are extracted using slicing operations and domain searches, and trajectory fitting is performed via Euclidean clustering. After obtaining a simple trajectory, an adaptive connection mechanism is introduced to enhance the algorithm’s practicality, thereby translating the algorithm’s intended outcomes into actual results. The proposed algorithm achieves a fitting accuracy exceeding 90%, with smoothness and average Z-direction values below 0.1, objectively demonstrating the high accuracy and stability of the trajectory fitting method presented herein. This study provides a feasible solution for automated welding of aviation Invar steel molds and offers new insights for the development of robotic welding trajectory planning.