Seismic Time-frequency Attribute And Application Based On Synchrosqueezing Wavelet Transform

Seismic attribute analysis is an important method for identifying oil and gas reservoirs.The seismic frequency band is limited to a certain range.Complex or thin-layer oil and gas reservoirs are difficult to distinguish.To improve the accuracy of gas reservoir identification and prediction,high-resolution methods are required.The signal time-frequency analysis technology can obtain the time-frequency information of the signal.Compared with the time-frequency analysis methods such as Fourier transform,continuous wavelet transform,and S transform,the frequency division processing based on synchrosqueezing wavelet transform has higher resolution and energy focus.This method is applied to the time-frequency analysis of seismic signals in oil and gas reservoirs,and the time-frequency attributes are extracted to improve the ability to identify and predict thin-layer oil and gas reservoirs.Synchrosqueezing wavelet transform is realized by rearranging complex wavelet coefficients in the frequency direction,and has high-resolution time-frequency characteristics.This paper analyzes the characteristics of synchrosqueezing wavelet transform by simulating seismic signals,which shows that synchrosqueezing wavelet transform has good reversibility and noise resistance.Integration parameter is an important variable of synchrosqueezing wavelet transform,which controls the reconstruction accuracy of signals with different signal-to-noise ratios.The integration parameter affects the reversible performance of synchrosqueezing wavelet transform,and the effects of the integration parameters of synchrosqueezing wavelet transform on the signal-to-noise ratio of reconstructed signals with different signal-to-noise are analyzed.Secondly,short-time Fourier transform,continuous wavelet transform,S transform and synchrosqueezing wavelet transform are applied to process simulation and real seismic data respectively,and time-frequency attributesare obtained.By comparing those of four methods,it is shown that synchrosqueezing wavelet transform has higher time-frequency resolution and energy focus.The method of realizing the frequency division property of seismic data is introduced,and the frequency division property is extracted by synchrosqueezing wavelet transform.This method is applied to model gas reservoir and an actual gas reservoir in Xinchang gas field.The results show that the time position and spatial distribution of gas bearing reservoir are identified and in accordance with that of reservoir beside the well.The advantages of the time-frequency attribute and application of the frequency division processing method based on synchrosqueezing wavelet transform are discussed.