铝合金MIG焊丝伸出长度视觉提取算法及动态过程辨识  

Vision-based identification model and extracting algorithm of wire extension in aluminum MIG alloy welding process

作　　者：石玗[1] 黄健康[2] 樊丁[1] 薛诚[2] 

机构地区：[1]兰州理工大学甘肃省有色金属新材料省部共建国家重点实验室,兰州730050 [2]兰州理工大学有色金属合金省部共建教育部重点实验室,兰州730050

出　　处：《焊接学报》2007年第8期1-4,共4页Transactions of The China Welding Institution

基　　金：国家自然科学基金资助项目(50675093);甘肃省自然科学基金资助项目(3ZS051-A25-029);教育部春晖计划资助项目

摘　　要：建立了一种基于视觉传感的铝合金MIG焊丝伸出长度检测系统,并提出了相应的图像处理算法,通过使用CCD摄像机,实时获取焊接区图像,根据焊丝特征定位其在图像中的纵坐标,然后对图像进行剪切,移位并叠加,使焊丝特征与背景之间的差别得到强化并提取焊丝伸出长度信息,结果表明,该算法快速有效。在以上工作的基础上,设计了阶跃响应试验,对焊丝伸出长度受送丝速度的影响规律进行了辨识,得到了数学模型,为铝合金MIG焊接过程中焊丝伸出长度传感与控制奠定了基础。A vision sensing system and corresponding image-processing algorithm for extracting the information of wire extension during aluminum MIG(metal inert-gas) welding process have been presented. The images of welding area are real-time captured by CCD camera. According the character of the wire in the image, its ordinate can be located by image processing algorithm. After image clipping, shifting and superposing, the difference between wire and background in the image will be enhanced, and then the information of wire extension can be extracted accurately. Experiments show that this method is effective and feasible. Based on the step response experimentation, the influence of wire speed on wire extension is identified and the mathematic model has been set. It provides the theoretical basis to realize the wire extension sensing and control in aluminum MIG welding process.

关 键 词：焊丝伸出长度 视觉传感 铝合金MIG焊 系统辨识 

分 类 号：TG409[金属学及工艺—焊接]