摘要
以采集的电阻点焊接头表面的数字图像作为信息源,探索了一种新的点焊质量无损监测方法。首先,通过图像特征分析,焊点表面图像被划分为4个环形特征区域,提取环形特征区域面积作为表征焊点质量的特征参数。其次,根据特征区域面积与焊点抗剪强度的相关性分析结果,选择了相关性显著的3个特征参数作为输入向量,焊点抗剪强度作为输出向量,建立了点焊质量的RBF神经网络监测模型。仿真分析和验证结果表明,基于焊点表面图像特征信息处理监测点焊质量的方法是可行的。
A new method was explored to monitoring joint quality based on information processing in digital image of welding spot surface in resistance spot welding.At first,through analyzing the image character,4 characteristic zones related to welding processing were mined from the image of welding spot surface.And then,their areas were measured to be taken as characteristic parameters for evaluating spot welded joint quality.Secondly,through the correlation analysis between 4 characteristic zones areas and tensile-shear strength of spot welded joint,3 characteristic parameters were selected as input vectors from them,and tensile-shear strength of the joint was target vectors.On the basis,Radical Basic Function neural network model was set up to estimate the weld quality.At last,the results of simulation and test show that it is feazible that spot-welded joint quality can be monitored based on image information of welding spot surface.
出处
《焊接学报》
EI
CAS
CSCD
北大核心
2006年第12期57-60,64,共5页
Transactions of The China Welding Institution
基金
国家自然科学基金资助项目(50275028)
关键词
电阻点焊
焊点表面图像
图像处理
RBF神经网络
质量监测
resistance spot welding
image of welding spot surface
image processing
Radical Basic Function neural network
quality monitoring
