Time-frequency analysis method based on improved matching pursuit algorithm

Abstract

The invention relates to a time-frequency analysis method based on an improved matching pursuit algorithm. The time-frequency analysis method comprises the steps of constructing a complex analysis signal of a seismic signal, and determining instantaneous phases and optimal master frequencies at the arrival time and the delayed time; calculating the master frequency and the phase of a time delay position, and determining an atom form according to the master frequency and the phase which are used as Ricker wavelet parameters; calculating an instantaneous amplitude of a wavelet under a wavelet energy normalization condition; matching an atomic amplitude according to a given percentage, and subtracting a matching wavelet which is in maximum correlation with the seismic signal from a seismic channel after four key parameters such as the atom arrival time, the atom master frequency, the atom phase and the atom amplitude are determined to obtain a residual signal, and carrying out recursion iteration on the residual signal within a certain threshold range to finish decomposition of original signal; stacking the time frequencies of all the matching wavelets to obtain a time frequency spectrum of the seismic signal. According to the time-frequency analysis method, the accuracy of atom master frequency matching is guaranteed under the condition that the calculation frequency of matching pursuit time frequency analysis is improved, so that the problem of discontinuity of an inphase axis on the section of a seismic frequency division body is solved.

Description

technical field [0001] The invention relates to the field of exploration geophysics, in particular to a time-frequency analysis method based on an improved matching pursuit algorithm in the field of petroleum geophysics exploration. Background technique [0002] The time-frequency analysis method of seismic data matching and tracking has many technical solutions for the matching of time-frequency atoms. The first technical solution is an adaptive search method based on the principle of maximum projection energy. The disadvantage of this method is that each sampling point must be When multiple parameters are scanned, a large number of inner product operations are performed in each iteration process, the calculation amount is too large, and the calculation time is huge. How to degrade the computational complexity of the matching pursuit algorithm and improve the computational efficiency of the matching pursuit algorithm is particularly critical. Other experts and scholars at ...

AI Technical Summary

Problems solved by technology

Other experts and scholars at home and abroad have also made many attempts in this area, introducing the genetic optimization algorithm into the atomic matching or introducing the instantaneous characteristics of the signal into the matching pursuit algorithm, which shortens the time-consuming, but improves the efficiency at the same time. Reduced accuracy of atomic matching and continuity of events in seismic crossover sections

Images