Scholarlink

<p>To improve the adaptability and diversity of welding robots for small- and medium-sized unstructured welding operations, this paper proposes a path planning method for welding robots. Firstly, the global vision sensor combined with the robot was used to obtain the point cloud from multiple poses and multiple angles to realize the global three-dimensional reconstruction of the weldment. The global semantic information of the weldment weld is recognized by the features; the global welding strategy planning of the welding robot is carried out, and it is guided to reach the initial welding position. They are aiming at the sudden change in tracking attitude that may occur when local laser tracking weld, a welding path segmentation method based on line laser tracking model, is proposed. Subsequently, the laser sensor is used to identify the precise welding path of the robot, and combined with the global attitude planning information, the welding posture of the robot is generated in real-time to complete the operation. Through tracking experiments, the results show that the robot path planning is reasonable and smooth, and the accuracy of weld trajectory extraction is high, which meets the production requirements of the welding process.</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

Path planning method of welding robot based on point cloud scene understanding

  • 期刊论文

  • 发表日期：

    2025年10月10日

查看全文
  • Jinyue Liu,
  • Yilong Li,
  • Xiaohui Jia,
  • Yusen Wu,
  • Tiejun Li

摘要

To improve the adaptability and diversity of welding robots for small- and medium-sized unstructured welding operations, this paper proposes a path planning method for welding robots. Firstly, the global vision sensor combined with the robot was used to obtain the point cloud from multiple poses and multiple angles to realize the global three-dimensional reconstruction of the weldment. The global semantic information of the weldment weld is recognized by the features; the global welding strategy planning of the welding robot is carried out, and it is guided to reach the initial welding position. They are aiming at the sudden change in tracking attitude that may occur when local laser tracking weld, a welding path segmentation method based on line laser tracking model, is proposed. Subsequently, the laser sensor is used to identify the precise welding path of the robot, and combined with the global attitude planning information, the welding posture of the robot is generated in real-time to complete the operation. Through tracking experiments, the results show that the robot path planning is reasonable and smooth, and the accuracy of weld trajectory extraction is high, which meets the production requirements of the welding process.