EMD算法在低温超导瞬变电磁噪声抑制中的应用  

EMD APPLICATION IN NOISE SUPPRESSION OFSUPERCONDUCTING TRANSIENT ELECTROMAGNETIC

作　　者：徐婷[1,2,3] 常凯[1,2,3] 王会武[1,2,3] 荣亮亮[1,2,3] 孔祥燕[1,2,3] 谢晓明[1,2,3] 

机构地区：[1]中国科学院上海微系统与信息技术研究所信息功能材料国家重点实验室 [2]中科院上海微系统所 [3]吉林大学"地球物理探测技术"联合实验室,上海200050

出　　处：《低温物理学报》2014年第5期401-404,共4页Low Temperature Physical Letters

摘　　要：瞬变电磁信号具有较宽的频谱,为使得超导瞬变电磁信号的接收不产生失真,要求接收信号的超导传感器具有足够高的带宽(100kHz以上),而接收高带宽信号会导致信号的噪声峰峰值变大,传统的滤波(模拟、数字)方法易造成衰减曲线畸变.基于数据驱动的经验模态分解(Empirical Mode Decomposition,EMD)算法能够根据信号本身特性对其进行信号分解和分量选择,具有自适应和不产生波形畸变的特点.本文采用EMD算法分别对数值模拟和野外实测的超导瞬变电磁信号进行EMD滤波处理和分析,结果表明,EMD可以显著压制瞬变电磁信号衰减曲线中的宽频带噪声且波形无畸变.经过EMD滤波的超导瞬变电磁与原始信号相比,信号衰减时间显著延长(从5ms延长至50ms),表明EMD算法能够极大提高瞬变电磁方法对地下深部信息的探测能力.In SQUID Transient Electro Magnetic, high bandwidth （more than 100 kHz） is essential, which also leads to larger noise peak. Traditional signal filtering methods may cause waveform distortion. We used EMD （Empirical Mode Decomposition, EMD） algorithm which decomposed signal into a series of IMF （Intrinsic Modal Function, IMF） ranging from high frequency to low frequency according to the features of the signal itself. Then we recombined part of the IMF and completed the filtering based on the signal＇s IMF energy distribution. In this work, we chose both simulation signal and real SQUID TEM signal from outdoor experiments to verify the feasibility and superiority of EMD. The result showed that large bandwidth noise was obviously suppressed without waveform distortion. Compared with the original signal, Attenuation was significantly longer （from 5ms to 50ms） which made it possible to detect the information of deeper underground.

关 键 词：SQUID 瞬变电磁 EMD IMF 

分 类 号：O511[理学—低温物理]