人机合作式管道漏磁信号分析与缺陷定量识别  

Man-Computer Signal Analysis for Oil-gas Pipe Magnetic Flux Leakage(MFL) and Inspection of Pipe Pit

作　　者：杨涛[1] 王太勇[1] 蒋奇[2] 

机构地区：[1]天津大学机械工程学院,天津300072 [2]山东大学控制科学与工程学院,济南250061

出　　处：《中国机械工程》2004年第6期488-490,共3页China Mechanical Engineering

基　　金：天津市自然科学基金资助项目 (993 80 2 411)

摘　　要：采用替换、平滑、反差等方法改善缺陷的漏磁信号 ,并把漏磁信号转变成云图 ,由人定性识别出缺陷的位置 ;通过对管道漏磁信号的分析 ,选取了缺陷漏磁场的特征量 ,采用非线性方法实现缺陷的定量分析。试验证明 。The paper present the methods of improving the signals of magnetic flux leakage for the pipes by means of replacing, smoothing and contrasting. The signals were transfered into the images and the places of the defects were confirmed by persons. The features were selected through analysing the signals of magnetic flux leakage. A nonlinear method was applied to complete quantitative recognition to defects. The experimental results show that the man-computer working method improves efficiency and precision.

关 键 词：管道 信号分析 人机合作 定量识别 特征量 

分 类 号：TE973.6[石油与天然气工程—石油机械设备] TG115.28[金属学及工艺—物理冶金]