基于NSTFT-WVD变换的芦山M_S7.0级地震前后地球天然脉冲电磁场信号时频分析

Time-frequency analysis of the Earth's natural pulse electromagnetic field signal before and after the Lushan M_S 7.0 earthquake based on NSTFT-WVD transform

本文针对地球天然脉冲电磁场（ENPEMF）信号的非平稳性特点,通过采用归一化STFT-WVD变换（NSTFT-WVD）提取芦山MS7.0地震前地表天然脉冲电磁场信号的时频特性,来研究大地震发生前地表天然脉冲磁场的时频特点。通过对线性调频信号分别进行Short-Time Fourier Transform（STFT）和WignerVille Distribution（WVD）,对比分析两者时频聚集的优缺点,提出采用NSTFT-WVD变换方法,可得到时频聚集特性更优且可抑制交叉项的时频分析方法。结果表明,对于ENPEMF非平稳信号,NSTFT-WVD变换能较为真实地反映震前ENPEMF信号的时-频-能量谱的联合分布特性,即震前1~2天呈现较为明显的脉冲全频率段静默状态,通道2和通道3数据的震前时频表现基本一致,具有较好的临震前兆的特点。

The present paper aims at the non-stationary characteristics of Earth＇s natural pulse electromagnetic field（ENPEMF）signals by using the normalized STFT-WVD（NSTFT-WVD）transformation,and the main analysis focuses on time-frequency characteristics of the ENPEMF signal before Lushan MS7.0earthquake.By comparing short-time Fourier transform（STFT）and Wigner-Ville distribution（WVD）transformation with the linear frequency modulation signal,this paper analyzes their advantages and disadvantages of both time and frequency focusing performance.NSTFT-WVD transform method can give us the better time-frequency aggregation and inhibits the cross-terms.The results show that NSTFT-WVD transform can reflect the real ENPEMF signal time-frequency-energy spectrum distribution before and after the earthquake,which could render more obvious silent state in entire frequency and sustained 1-2days before the earthquake.The timefrequency representation of data channels 2and 3are basically consistent with this characteristic,which well represents the feature of impending earthquake precursors.