Full Waveform Inversion Method in Time Domain Based on Zero-Mean Normalized Cross-Correlation Objective Function

Abstract

The invention relates to a time domain full-waveform inversion method based on zero-mean normalized cross-correlation objective function. Calculation of zero-mean normalized cross-correlation betweensimulation and observation records is used for replacing calculation of a least square objective function in the traditional full-waveform inversion. On the basis of the normalized cross-correlation,absolute values of amplitudes of the simulated data and the observed data are kept to be between -1 and 1, so that the inversion result error caused by the amplitude error is reduced effectively. Because of the normalization of the amplitude energy, the energy weight of the offset in the seismic record is improved; and since the offset data carry information of the large scale structure in the model, the occurrence of the cycle skipping is reduced. Because the actual seismic record is easy to affect by the low frequency noises and the noises are non-mean noises, a de-mean item is introduced; and the interference of low frequency noises is reduced by de-mean value calculation of the simulation and observation records. With the method provided by the invention, the velocity parameter of theunderground medium is inverted accurately under the circumstance of being lack of the low frequency information or noise pollution on the low frequency information.

Description

technical field [0001] The invention belongs to the technical field of seismic exploration, and relates to an underground medium imaging and medium velocity calculation method in the field of seismic exploration, in particular to a time-domain full waveform inversion method, in particular to a zero-mean normalized cross-correlation objective function The time domain full waveform inversion method. Background technique [0002] Non-renewable resources such as petroleum and minerals play a vital role in all aspects of human life. Therefore, the corresponding resource exploration technology is constantly innovating and developing. In recent years, with the increasing scale of land and offshore oil and gas development, exploration targets have changed from shallow simple structural geological areas to deep complex structural areas and hidden oil and gas reservoirs, and the difficulty of exploration has increased. In order to further solve the problem of deep energy exploration...

AI Technical Summary

Problems solved by technology

[0004] The main problem faced by the full waveform inversion method is the cycle skip problem, that is, the waveform phase difference between the simulated record and the observed record is more than half a cycle, resulting in a waveform matching error, which leads to an error in the inversion result

There are two main reasons for the cycle skipping. One is that the initial model of the inversion is too far from the real underground model, resulting in a large difference between the simulated record waveform and the observed record waveform, which leads to the cycle skipping; the other is the lack of low-frequency information, the seismic records The low-frequency signal in the medium carries the information of the large-scale structure and is not easy to skip cycles

However, low-frequency information is often missing in actual observation records, and there are many reasons for the lack of low-frequency information, such as missing low-frequency information from seismic wavelets excited by the source, low-frequency information not being received by the geophone, or low-frequency information being artificially removed due to noise pollution, etc.

For the traditional full waveform inversion based on the least squares objective function, the requirements for amplitude information are relatively high, and the waveform amplitude error caused by source excitation or seismic record acquisition has a great impact on the inversion process

Images