Study And Practical Application Of Time-frequency Analysis Method For Seismic Signals

The seismic data can be used to invert the dispersion characteristics related to oil and gas.At present,the technology mainly uses the pre-stack seismic data for inversion,which is called frequency-dependent AVO inversion technology.It is a research hotspot in geophysical exploration technology.The key technology in the inversion is to obtain the information of different frequency components by time-frequency analysis of pre-stack seismic data,and then combine the AVO inversion technique to obtain the characteristic parameters related to the seismic wave velocity dispersion.Therefore,when time-frequency analysis of actual seismic data is performed,it is crucial to select an appropriate time-frequency analysis method.Time-frequency analysis of signals is a very important area of signal processing,which can reveal more detailed information in the data.The actual seismic data processing is faced with non-stationary signals.The time-frequency joint analysis is to obtain the time-frequency domain joint information of the original signal,and to characterize the time-varying spectral characteristics of the original signal.The paper mainly analyzes five d ifferent time-frequency analysis methods,including short-time Fourier transform,continuous wavelet transform,generalized S transform,smoothing pseudo Wigner-Ville distribution and Hilbert-Huang transform.Different methods are studied from the traditio nal linear time-frequency analysis and high-precision time-frequency analysis.The time-frequency analysis and comparison analysis are completed by synthetic seismic records,and the validity and feasibility of these time-frequency analysis methods are verified.At the same time,the paper also discusses how to select appropriate,reasonable and effective parameters for different time-frequency analysis methods.By comparison,it is found that the smoothing pseudo Wigner-Ville distribution time-frequency analysis technology has the characteristics of high computational efficiency,high time-frequency aggregation and strong ability to remove cross-interference items.A smoothing pseudo Wigner-Ville distribution time-frequency analysis technique suitable for actual seismic data processing is formed through research and development,and the time-frequency analysis method is applied to the frequency-dependent AVO inversion of actual seismic data.In this paper,the seismic data of an actual work area is selected,and the application research of pre-stack frequency-dependent AVO inversion is carried out.In this paper,the time-frequency analysis of the actual pre-stack seismic data is performed by using the time-frequency analysis method of the smoothing pseudo Wigner-Ville distribution,and the frequency division processing of the actual seismic data is completed.Finally,the frequency-dependent AVO inversion method is used to obtain the seismic dispersion characteristic attribute parameters,and the gas abundance prediction based on the dispersion characteristic parameter is completed.