The Applications Of Time-frequency Analysis Methods With High Resolution In Reservoir Prediction

As Fourier transform reflects global information, it cannot show time frequency local characteristics of the signal. Conventional Fourier transform is only applicable to stationary signal analysis, is not suitable for non-stationary signal. Seismic signals are non-stationary signals. Analyses that are only in time domain or frequency domain cannot meet the needs of practical application; therefore, we should use the signal time-frequency analysis methods to show the characteristics of the seismic signal.The time-frequency analysis methods have been a hot topic in signal processing domain in recent years. Using time-frequency analysis methods can show association characteristics of time and spectrum frequency in time-frequency distribution graph.This paper introduces the short-time Fourier transform, wavelet transform and S transform, generalized S transform, deconvolution short time Fourier transform and time-frequency analysis method based on the matching pursuit t and their advantages and disadvantages.In recent years, the reservoir prediction using low frequency information of seismic signals has made great progress. More and more people begin to pay attention to the applications of low frequency information. Theories and practices show that when seismic waves go through the fluid, the high frequency part energy of seismic signals are absorbed. Low-frequency shadow can be used as a direct indicator of oiliness.In the letter, computer simulations of several theory signals using several time-frequency analysis methods show that this method achieves better results compared with some traditional time-frequency representation methods. The method not only improves the time-frequency resolution but also reduces the cross-terms.Based on the theory of low-frequency shadows, we use the deconvolutive Short-Time Fourier Transform(DSTFT) spectrogram method to do the spectral decomposition and detect oil-gas potential of the reservoirs. Real data processing proves the feasibility of the method to detect oil and gas.