大地电磁信号具有频带范围宽、信号微弱的特点,在城市进行信号采集时,极易受到人文电磁噪声的干扰使信噪比下降,导致反演结果出现偏差.传统Hilbert-Huang变换去噪中经验模态(Empirical Mode Decomposition,EMD)分解过程中极易出现模态混叠现象,笔者提出基于变分模态分解改进的 Hilbert-Huang变换(improved Hilbert-Huang transform based on Varational Mode Decomposition,VMD-HHT)方法对城市强人文干扰进行压制去噪.结果表明:改进的Hilbert-Huang变换能够有效避免模态混叠现象,明显改善电磁数据的信噪比,去噪后的数据可为后续视电阻率反演提供高质量的时间序列.
Application of VMD-HHT in magnetotelluric strong human interference denoising
The magnetotelluric data usually has a wide frequency range and weak target signal.When the signal is collected in the city,it is very easy straightforward to be interfered with by the human electromagnetic noise,which reduces the signal-to-noise ratio,and results in the deviation of the inversion results.Mode aliasing often occurs in the process of empirical mode decomposition(EMD)in the traditional Hilbert-Huang transform denoising.In this paper,an improved Hilbert-Huang trans-form method based on Variational Mode Decomposition(VMD-HHT)is proposed to suppress and denoise the strong human e-lectromagnetic noise in cities.The results show that the improved Hilbert-Huang transform can effectively suppress modal aliasing,and significantly improve the signal-to-noise ratio of electromagnetic data.Thus it can provide high-quality time series for subsequent apparent resistivity inversion.
国家科技期刊平台
NSTL
万方数据
