基于盲源分离和时频分析的漏磁信号处理  

Study on magnetic flux signals process based on blind source separation and timefrequency analysis

作　　者：胡浪涛[1] 何辅云[2] 查君君[3] 

机构地区：[1]安庆师范学院 [2]合肥工业大学 [3]安徽工程科技学院

出　　处：《电子技术（上海）》2008年第1期90-93,共4页Electronic Technology

基　　金：国家科技部科研院社会公益研究资金项目(Z00-G03);安徽省"十五"科技攻关资助项目(040510E2)

摘　　要：利用独立分量分析的冗余取消特性,对多维加噪声观测信号进行盲源分离,得到源观测信号,实现噪声的有效消除,文章中应用此方法处理了仿真漏磁缺陷信号,实验结果表明:该除噪方法能极大提高漏磁信号的信噪比,且其效果要优于小波变换除噪方法。漏磁缺陷信号在时域上波形非常相似,很难加以分辨,而它们的危害性却大不相同,文章对裂纹和凹坑缺陷信号进行小波包分解,根据信号特征自适应的产生一组最优基来表征信号,分析这两种缺陷的时频特性,准确识别出这两种缺陷。In order to remove noise from observed signal,a new denoising method based on the redundancy reduction capability of the independent component analysis(ICA)was Proposed,the blind source separation(BSS)of ICA was applied to the extended observed signal,thus the noise embedded in the observed signal was removed,This method has been simulated in Magnetic flux leakage signal. Experimental results show that signal noise ratio of magnetic flux signal can be improved greatly and ICA algorithms are superior to wavelet denoising method,the magnetic flux leakage defect signals waveforms in the time domain are very similar,we can't differ them,but the harmfulness is different, we decomposed some pipeline MFL signals(the cracks and pockmarks signals),to adjustably produce a best base as a token to signal according to the different signal,Analysed time-frequency characteristics of two defects,found out their feature to identify them.

关 键 词：漏磁 独立分量分析 小波包 时频分析 

分 类 号：TN911[电子电信—通信与信息系统]