Research On High Resolution Seismic Signal Processing Method Based On Variational Modal Decomposition

Improving the resolution of seismic data is a key step in seismic signal processing,as well as,it is also an important link in the exploitation of oil,natural gas and other resources.However,as the mining work enters the deep layer from shallow to deep,the geological conditions become more and more complex,which makes the seismic reflection weak,the phase of the same phase is discontinuous,the oil-bearing response is weaker,and the attenuation is lager,resulting in extremely low resolution of the acquired seismic data.Therefore,we need to further study the method of improving the resolution of seismic data.Based on the time-frequency analysis of Variational Mode Decomposition(VMD),combined with spectral whitening and wavelet threshold de-noising,this paper studies the method of improving seismic data resolution in high-frequency compensation and high-frequency noise suppression of seismic data.Then,the main research contents are as follows:(1)Based on the research of traditional time-frequency analysis method and time-frequency method based on variational mode decomposition,we have compared and analyzed the short-time Fourier transform,wavelet transform,time-frequency analysis method based on empirical mode decomposition and variation-based.Then,this paper verifies the high time-frequency resolution characteristics of the time-frequency analysis method based on variational mode decomposition.(2)Spectral whitening method based on time and frequency domain were analyzed,we have known that based on frequency domain is proved fast and simplicity.In the time-frequency analysis based on variational mode decomposition method,on the basis of combing spectral whitening method based on frequency domain,this paper proposes new seismic signal processing methods of high resolution,and the spectral whitening method based on Hilbert-Huang transform were analyzed,verified the spectral whitening method based on variational mode decomposition has better seismic data resolution.(3)Based on the study of traditional seismic data amplitude compensation method,a high-resolution seismic signal processing method based on time-frequency analysis and amplitude compensation method is proposed.Then,compared with the method based on empirical mode decomposition,the high time domain resolution of the amplitude compensation method based on variational mode decomposition is verified.(4)Having studied the wavelet threshold method,we analyzed the characteristics of suppress noise about wavelet threshold de-noising method.Then,A new seismic signal noise suppression method is proposed.and,we also compared with the traditional wavelet threshold de-noising method.We known that the wavelet threshold de-noising method based on variational mode decomposition is proved to have better noise suppression ability.Through synthetic seismic record processing and actual seismic data application,the results show that the spectral whitening method based on variational mode decomposition,the amplitude compensation method based on VMD and the wavelet threshold de-noising method based on VMD can effectively improve the resolution of seismic signals.And,compared with the proposed spectral whitening method based on HHT,amplitude compensation method based on EMD and the traditional wavelet threshold de-noising method,the proposed method is quite better.