Time-frequency domain full-waveform inversion method and device using normalized seismic sources

Abstract

The invention discloses a time-frequency domain full-waveform inversion method and device using normalized seismic sources, the seismic sources are a plurality of single-frequency signals extracted from known wavelets, the amplitudes of the single-frequency signals are normalized, and wave field simulation of the extracted normalized single-frequency seismic sources is completed in a time domain; and a phase-sensitive detection method is adopted to decouple the forward propagation wave field and the backward propagation wave field respectively, so that amplitude and phase information of the wave fields is obtained through decoupling, limitation of a conventional band-limited wavelet amplitude spectrum on each frequency amplitude in the inversion process is eliminated, and then a gradient calculation expression is calculated according to the amplitude and phase information of the wave fields and the gradient calculation expression. The gradient of the velocity field model parameters is obtained through calculation, so that the accurate updating direction of the velocity field model parameters can be obtained, and the purpose of simultaneous inversion of multiple frequencies in inversion is achieved. According to the method, the limitation of the neutron amplitude spectrum on the amplitude of the single-frequency signal in the conventional inversion can be weakened, the wavelet effect in the inversion is eliminated, the inversion result of spectrum whitening is realized, and finally, the high-precision inversion result can be obtained.

Description

Technical field [0001] The present invention relates to a time-frequency domain full waveform inversion method and apparatus using normalized sources, belonging to the field of seismic exploration speed modeling technology. Background technique [0002] With the petroleum exploration to deep and ultra-deep metastasis, the high-precision imaging problem of deep target layer high-side complex structure has become the current constraints of oil exploration and development is imminent. Speed ​​modeling is a soul that achieves complex construction imaging, and traditional speed modeling techniques such as speed chromatography have achieved great success in the past decades, but it has not been able to meet the needs of high-resolution seismic imaging. Therefore, it is urgent to develop a new generation of speed modeling techniques to meet the needs of complex construction imaging. [0003] Currently, Full Waveform Inversion (FWI) is a recognized most potential new generation speed mod...

AI Technical Summary

Problems solved by technology

However, due to the needs of data simulation, the frequency contained in the wavelet has a certain amplitude distribution, that is, only the frequency update near the main frequency is dominant, and the update of more low-frequency information cannot be guaranteed; at the same time, it is impossible to calculate the gradient. Decoupling and normalizing the frequency features, so there is a wavelet effect in the full waveform inversion, and there will be a phenomenon of wrong update at the structural interface, which reduces the inversion accuracy to a certain extent; even if the multi-scale method is used, this subwave Wave effects also accompanies the entire inversion, leading to over-updating at the high-contrast boundaries of structures

Images