The Sparse Inversion Spectral Decomposition And Its Application

The time frequency analysis is a fundamental tool in seismic processing,and the variation of frequency content with time in the seismic signal can be observed by this method.These abnormal variations can be used to identify the geologic structure and oil and gas reservoir.The sparse inversion spectral decomposition(SISD)is an unconventional time frequency analysis method and it has higher time frequency resolution than conventional methods.The inversion equation is built based on the convolution model,then the L1 norm is chosen as the sparse constraint to build the objective function and the fast iterative soft thresholding algorithm is used to obtain the optimal solution.In this paper,the influence of the wavelet dictionary and regularization parameter on the time frequency spectrum is discussed.Then,the 1D synthetic signal is used to verify the high resolution,noise immunity and the identification of the phase information about the signal.At last,the SISD is applied on the real data,including the identification of faults,channel complex and gas reservoir:(1)The obscure faults are found by taking use of the high resolution phase spectrum.(2)Both of the amplitude spectrum and phase spectrum are fully used to detect the channel complex,and some small scale channels are successfully to detected.Furthermore,the phase spectrum shows great ability to identify the boundary of channel complex.(3)The low frequency shallow and time frequency attribute are utilized to successfully find the gas reservoir.The SISD is a high resolution time frequency method and it has great potential in seismic exploration.