Hilbert变换在钢管漏磁检测缺陷定量化识别中的应用

提出一种根据漏磁检测信号切向分量信息构造法向分量的新方法.根据Hilbert变换原理,分析了钢管缺陷漏磁信号各个分量之间的正交对偶关系,论证了缺陷漏磁信号各个分量之间相互转换的理论基础和条件.基于探头所检测到的缺陷信号切向分量,利用Hilbert变换构造法向分量,增加了定量化识别的特征量,改善了缺陷识别的准确率和尺寸定量化误差.理论和仿真结果证明了该方法的有效性,可实现一维探头的准二维测量.

Quantitative Analysis of Magnetic Flux Leakage Signal for Steel Pipe Defects Using Hilbert Transform

A novel method is proposed using Hilbert transform to construct normal component complying with tangent component of magnetic flux leakage testing (MFLT) signals. Hilbert transform is presented from the mathematics definition at first in the paper, and then the basic principle and process of discrete Hilbert transform are discussed. According to the theory that three components of MFLT signals are orthogonal each other, the method for transforming from one component to another is proposed and demonstrated. Application example of actual MFLT signals demonstrates that the proposed method can be used to extract more feature quantity, and hence improves the accuracy rate of defect identification as well as quantitative analysis evidently.