A Simultaneous Inversion Method for Reflection Coefficient Sequences of Multi-channel Seismic Records

Abstract

The invention discloses a multichannel seismic record reflection coefficient sequence simultaneous inversion method. The method thoroughly considers sparse characteristics of a formation reflection coefficient sequence and spatial transverse continuity of seismic channels, a parameterized exponential function is used to measure the sparsity of the reflection coefficient sequence, and a corresponding constrained optimization objective function is built. Firstly, based on the characteristics of a convolution model and the sparsity hypothesis of the reflection coefficients, on the premise of notlosing any information, a dimension reduction preprocessing matrix is designed, and the dimension of the constrained objective function is reduced; and then, a trust region sequential quadratic programming algorithm is used for high-efficiency solution. The algorithm has global convergence and a super linear convergence rate, the calculation efficiency is high, and the inversion precision is high.Compared with the traditional single-channel seismic record deconvolution algorithm, the algorithm of the invention has a more stable and reliable inversion result as the spatial correlation of the seismic channels is considered.

Description

technical field [0001] The invention belongs to the technical field of geophysical exploration, and relates to a seismic record reflection coefficient sequence inversion method, in particular to a multi-channel seismic record reflection coefficient sequence simultaneous inversion method. Background technique [0002] In stratigraphic-lithological reservoirs, thin interbedded reservoirs are one of the main reservoir types. Due to the limited frequency range of the artificial seismic source used to stimulate the earthquake, the travel time difference of the reflected waves generated by the thin layer is small, so that the reflected waves superimpose and interfere with each other, and the received seismic records cannot directly distinguish the thin interbedded structure, which is the identification of thin interbedded reservoirs. bring about many difficulties. Therefore, it is necessary to improve the resolution of reflection seismic records to achieve high-precision characte...

AI Technical Summary

Problems solved by technology

[0006] (1) Conventional reflection coefficient sequence inversion algorithms are usually based on gradient or conjugate gradient to construct the descent direction, and its convergence rate is linear and the convergence speed is slow;

[0007] (2) The conventional inversion of reflection coefficient series is usually performed trace by trace, ignoring the lateral continuity of seismic records, which makes the low-frequency noise of subsequent stacking impedance serious

Images