摘要
转向架作为列车走行装置,其性能直接影响着列车运行舒适性以及安全可靠性。构架作为转向架骨架,在转向架机构中起着举足轻重的作用。CRH系列转向架构架均为钢焊接结构,而钢焊接结构易出现裂纹、漏焊、焊缝外观不规则等缺陷。采用CRH1转向架构架焊接工艺中使用较多的T型焊接,使用PCA降维处理得到相应的"特征焊缝"向量,利用MATLAB GUI实时抓拍焊缝图像并提取焊缝与"特征焊缝"对比,基于设定的阈值进行焊缝的合理性判断,并给出报警提示。经过大量的焊缝采集图像测试,证明采用焊缝图像检测在转向架构架焊接表面质量检测中的应用是合理的。
Bogie is traing running gear and its safety performance affects operation comfort, the safety and reliability of trains. Bogie frame, as the skeleton, plays a pivotal role in the bogie system. For the current CRH bogie frames, frames are all steel welded structures, which usually induce some faults, like crack, lack of weld, irregular weld shape, etc. In this paper, based on T welds of CRH1 bogie flame weld assembly process, PCA dimension reduction processing algorithm was used to calculate and gain ＇features weld＇ vector. The system used MATLAB GUI software to capture real -time image and extract the weld seam, which were then compared with ＂ features weld; to judge the rationality of the weld seam captured and gave alarm prompt by means of setting the range of threshold value. Through a large number of weld images tests, it proves that image recognition technology is proper and reasonable in the application of bogie flame welding quality detection.
出处
《铁道科学与工程学报》
CAS
CSCD
北大核心
2015年第2期402-406,共5页
Journal of Railway Science and Engineering
基金
国家自然科学基金重点项目(61134002)
关键词
转向架构架
T型焊接
焊缝表面质量检测
图像识别
bogie frame
T- joint fillet weld
weld surface quality detection
image recognition
