Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
6867689 | Robotics and Computer-Integrated Manufacturing | 2019 | 13 Pages |
Abstract
In welding, the actual weld seam is the target path of welding robot. When the deviation between the actual pipelines and the ideal models cannot be ignored, special research should be carried out. This paper introduces a novel method to quantify and compensate the path deviation of actual intersecting curve, which is used to determine the welding path of robot. Weld location is a technical means, which can be used to find the position of some key points on intersecting curve. Combining the information of weld location with the inherent characteristics of intersecting curve, this paper has analyzed the results of actual welding experiments, all these works laid the foundation for the proposed algorithm. First of all, on the basis of previous studies, this paper introduces a kind of geometric model for ideal intersecting pipelines, this model can cover most of the ways of intersection. Secondly, based on this model, the factors leading to the deviation between the theoretical intersecting curve and the actual are analyzed. Generally speaking, there may be some connections or coupling between the deviation sources. In view of the above problem, this paper constructs model for each source of deviation and discusses the relationship between them, then quantifies the main sources of deviation by using the weld location information, especially the quantification of the main pipe's ovality. Subsequently, the deviation after the quantification is compensated to the parametric equations of the theoretical intersecting curve. As the goal of this paper, the actual intersecting curve's equations based on non-ideal model are obtained, which is regarded as the path of welding robot. The correctness and flexibility of the path planning approach are verified by the welding robot at last.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Liu Yan, Liu Jiang, Tian Xincheng,