天然气管道漏磁检测中的信号处理

采用漏磁法对天然气管道进行检测时，关键问题是对漏磁信号的处理。为此，介绍了去除奇异值的算法，用相邻两点的平均值代替奇异值，可去除脉冲干扰而对信号没有影响。对于漏磁信号中的高斯噪声，采用小波变换去噪方法，选取二阶样条小波为小波函数，选用软阈值函数和固定阈值的方法处理小波系数。提出了提取缺陷信号特征的梯度算法，用信号幅度和梯度值的变化作为缺陷估计标准。研究结果表明，漏磁信号被处理后，很好地去除了各种信号噪声，有助于提取出缺陷信号的特征，可实现对天然气管道的无损检测。

SIGNAL PROCESSING OF MAGNETIC FLUX LEAKAGE TESTING FOR GAS PIPELINE 1)

When detecting gas pipelines by magnetic flux leakage method, the key is how to process the signals of magnetic flux leakage. Here the algorithm of extremum filtering is introduced. Substituting the extremum with the average of the adjacent two points, the pulse noise can be filtered without any effect on the signals. As for the Gaussian noise in the signals of magnetic flux leakage, the wavelet based de-noising technique is used taking the second-order spline wavelet as wavelet function and processing the wavelet coefficients by soft threshold function and fixed threshold value. The gradient algorism is developed to extract the defect signal. The variation of signal amplitude and gradient is used as the standards of the defect assessment. The study results show after the signals of magnetic flux leakage are processed, all kinds of noise can be eliminated well. It is helpful to extract the features of defect signals for nondestructive detection of gas pipeline.