摘要
不同形状、材质、结构的管道附件对超声波能量衰减和传播路径的影响不同,准确识别超声信号所表征的管壁特征变化,针对不同管件进行超声信号特征量分析,可深入探索复杂管件结构超声成像机理。综合利用A扫描波形、B扫描图像、C扫描图像进行缺陷多维度分析,能够有效完成内外缺陷识别和缺陷精细分类。通过提取管件及其缺陷的灵敏维度特征,可实现海量超声检测数据管件智能识别与缺陷自动量化,提高超声检测精度,降低误判率。
Pipe accessories with different shapes,materials and structures have different influences on ultrasonic energy attenuation and propagation path.Accurate identification of the changes of pipe wall features characterized by ultrasonic signals and analysis of ultrasonic signal feature for different pipe fittings can deeply explore the ultrasonic imaging mechanism of complex pipe fittings.Comprehensive use of A-scan waveform,B-scan image and C-scan image for multi-dimensional defect analysis can effectively complete internal and external defect recognition and fine defect classification.By extracting sensitive factors reflecting the structure of pipe fittings and defect characteristics,this paper realizes intelligent identification of pipe fittings and automatic quantification of defects with massive ultrasonic testing data,improves the accuracy of ultrasonic testing and reduces the misjudgment rate.
作者
王怀江
孙晨号
张天俭
刘帅
徐枭
WANG Huaijiang;SUN Chenhao;ZHANG Tianjian;LIU Shuai;XU Xiao(China National Offshore Oil Corporation(Tianjin)Pipeline Engineering Technology Co.,Ltd.,Tianjin 300452,China)
出处
《兵器装备工程学报》
CAS
CSCD
北大核心
2022年第S02期375-379,共5页
Journal of Ordnance Equipment Engineering
关键词
输油管道
超声检测
特征识别
数据分析
oil pipeline
ultrasonic testing
feature recognition
data analysis
