基于磁控旋转电弧传感的堆焊层曲面重建关键技术研究

发布时间：2018-05-10 07:32

本文选题：磁控旋转堆焊传感器 + 焊道自动排布　；参考：《湘潭大学》2017年硕士论文

【摘要】：焊接自动化技术是工业技术快速发展的推动因素之一,世界各国对工业技术的重视程度不断加大,为焊接自动化技术的发展揭开了新的篇章。堆焊是一种常用的零件修复和材料表面改性技术,具有广泛的应用前景,但是现阶段堆焊的自动化程度较低,整体质量不高,缺乏一种能有效提高堆焊生产过程自动化程度的方法。本文采用磁控旋转堆焊传感器进行数据采样,对相关技术理论进行了研究,解决了堆焊层曲面重建的关键性问题,提出一种基于磁控旋转电弧传感的堆焊层曲面重建方法,用于焊道排布偏差识别与控制,实现焊道的自动排布,具体内容如下:(1)设计出一种满足堆焊层焊道坡口采样要求的磁控旋转堆焊传感器。通过对电弧在可控磁场内的运动形态进行分析,确定了磁控电弧旋转的采样方式。针对堆焊条件下对焊接传感器的要求,经有限元磁场仿真分析,设计出一种可用于堆焊层曲面特征信息采集的磁控旋转堆焊传感器。(2)求解焊道坡口的弧长数学模型,仿真分析不同偏差时信号的变化情况。通过分析磁控旋转堆焊传感器的采样机理和堆焊不同焊道排布策略对电弧形态的影响作用,建立弧长数学模型,再根据磁控旋转堆焊系统传递模型,对不同偏差情况的电流信号进行仿真分析,总结出堆焊层焊道坡口形貌变化与电流信号的对应关系。(3)提出一种基于采样周期的分区间增长Delaunay三角剖分的堆焊层曲面重建方法,重建堆焊层焊道坡口的三维曲面模型,进行焊道排布偏差识别与控制。根据磁控旋转堆焊传感器的传感模型和弧长数学模型,提取出采样数据点的信息,借助MATLAB软件的科学计算和可视化图形分析功能,将采样数据点按采样周期不同进行分区间Delaunay三角剖分,建立堆焊层的三角网格模型,加入等效高度模型后进行堆焊层曲面重建。通过改进Kriging插值算法平滑处理,解决了曲面模型细节特征表述模糊和平滑度不足的问题,再通过上位机调用GPU内核并行运算解决了实时性低的问题,重建堆焊层焊道坡口的三维曲面模型,用于堆焊焊道排布的偏差识别与控制,实现焊道的自动排布。(4)通过进行磁控旋转电弧堆焊试验,验证堆焊焊道的自动排布效果。根据提出的堆焊层曲面重建方法,搭建磁控堆焊试验平台进行焊接试验,结果表明本文所述方法焊道排布偏差识精度高,焊道自动排布效果好。
[Abstract]:Welding automation technology is one of the driving factors for the rapid development of industrial technology. The importance of industrial technology is increasing in the world, which opens a new chapter for the development of welding automation technology. Surfacing welding is a commonly used part repair and material surface modification technology, which has a wide application prospect, but at present, the automation degree of surfacing welding is low, and the overall quality is not high. There is a lack of an effective method to improve the automation of surfacing production process. In this paper, the magnetic control rotating surfacing sensor is used to sample the data, and the related technology theory is studied. The key problem of surface reconstruction of surfacing layer is solved, and a method of surface reconstruction of surfacing layer based on magnetic control rotating arc sensor is proposed. It is used to identify and control the distribution deviation of welding pipe and to realize the automatic layout of welding pipe. The concrete contents are as follows: 1) A magnetic control rotating surfacing welding sensor is designed to meet the requirements of groove sampling in surfacing layer. The sampling mode of the arc rotation is determined by analyzing the movement of the arc in the controlled magnetic field. According to the requirement of welding sensor under surfacing condition, a magnetic control rotating surfacing sensor. 2) which can be used to collect the characteristic information of surfacing layer surface, is designed by finite element magnetic field simulation analysis to solve the arc length mathematical model of welding path groove. The variation of signals with different deviations is analyzed by simulation. By analyzing the sampling mechanism of the magnetic control rotating surfacing welding sensor and the influence of different laying strategies of the surfacing welding on the arc shape, the mathematical model of arc length is established, and then according to the transfer model of the magnetic control rotating surfacing welding system, the mathematical model of arc length is established. Based on the simulation analysis of the current signals with different deviations, the corresponding relationship between the groove shape change and the current signal of the surfacing welding layer is summarized. A surfacing layer surface reconstruction method based on sampling period is proposed, which is based on the interval growth Delaunay triangulation of the surfacing layer. The 3D curved surface model of the bevel of surfacing layer was reconstructed to identify and control the distribution deviation. According to the sensor model and arc length mathematical model of magnetically controlled rotating surfacing welding sensor, the information of sampling data points is extracted, and the functions of scientific calculation and visual graphic analysis of MATLAB software are used. The sampling data points are divided into Delaunay triangulation according to different sampling periods. The triangular mesh model of surfacing layer is established and the surface of surfacing layer is reconstructed by adding equivalent height model. By improving the smoothing of Kriging interpolation algorithm, the problems of fuzzy representation and lack of smoothness of the detailed features of the surface model are solved, and the problem of low real-time performance is solved by calling the GPU kernel parallel operation on the host computer. The 3D curved surface model of the bevel of surfacing layer is reconstructed, which is used to identify and control the deviation of the surfacing welding pipe placement, and to realize the automatic placement of the welding pipe. The magnetic controlled rotating arc surfacing test is carried out to verify the effect of the automatic placement of the surfacing welding pipe. According to the surface reconstruction method of surfacing layer, a magnetic controlled surfacing test platform is built for welding test. The results show that the method described in this paper has high recognition accuracy and good automatic placement effect.
【学位授予单位】：湘潭大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG455

【参考文献】