矿集区采集的大地电磁信号极易受到各类噪声污染,导致其视电阻率-相位曲线在低频段出现紊乱现象或呈现出近源效应等.文中提出了一种优化固有时间尺度分解(Improved Intrinsic Time Decomposi-tion,IITD)和小波阈值(Wavelet Threshold,WT)的大地电磁(Magnetotelluric,MT)去噪方法及应用.首先将含噪信号进行IITD分解得到若干阶旋转(Proper Rotation,PR)分量;然后对PR分量进行小波去噪,叠加小波系数重构得到MT去噪数据.通过计算机模拟出不同类型的强噪声,并对小波阈值法设置不同的分解层数、基函数对强噪声进行处理,总结出该算法面对不同噪声时的去噪性能.对模拟大尺度方波和三角波噪声去噪后,信噪比最高可达24dB和17dB,所提方法去噪性能显著.将所提方法应用至MT实测数据的降噪,结果显示该方法能够有效去除隐藏在MT数据中的强噪声.由去噪前后视电阻率曲线对比可知,相较于远参考法和原始曲线,所提方法获得的视电阻率曲线更为光滑、连续,低频段的数据质量明显改善.
Magnetotelluric Noise Suppression Method Based on IITD-WT
The magnetotelluric signals collected in the mining concentration area are easily polluted by various types of noise,resulting in the disorder of their apparent resistivity-phase curves in the low-frequency band or showing near-source effects,etc.We propose a magne-totelluric(MT)denoising method to improve intrinsic time scale decomposition(IITD)and wavelet thresholding(WT)and their applications.The noise-containing signal is firstly de-composed by IITD to obtain a number of order rotational(PR)components;then the PR component is denoised by wavelet,and the wavelet coefficients are superimposed to recon-struct the MT denoised data.Different types of strong noise are simulated by computer,and different decomposition layers and basis functions are set for WT method to remove the strong noise.The denoising performance of the algorithm in the face of different noises is summarized.After denoising the simulated large-scale square wave and triangular wave noise,the signal-to-noise ratio can reach up to 24 dB and 17 dB,and the proposed method has significant denoising performance.The proposed method can be applied to the noise re-duction of MT measurement data,and the results show that the method can effectively re-move the strong noise hidden in the MT data.From the comparison of the apparent resistivi-ty curves before and after denoising,it can be seen that the apparent resistivity curves ob-tained with the proposed method are more smooth and more continuous,and the data quality in the lower frequency bands is higher compared with the far-reference method and the origi-nal curves.
magnetotelluricimproved intrinsic time scale decompositionwavelet thresholddenoising
万方数据
维普
