论文详情

基于LSQR算法的谱反演方法研究

石油物探

2012年 51卷第No. 1期

阅读：72

查看详情

Title

Spectral inversion method analysis based on the LSQR algorithm

单位

(1.中国石油大学（华东）地球科学与技术学院,山东青岛266555;2.中国石油化工股份有限公司胜利油田分公司,山东东营257000;3.北京诺克斯达石油科技有限公司,北京100192)

Organization

Chai Xintao,School of Geosciences,China University of Petroleum(East China),Qingdao 266555, China

摘要

阐述了谱反演原理,给出了谱反演流程。通过单道模型、楔形模型和Marmousi模型的应用分析,验证了谱反演从地震记录中去除地震子波影响进而得到反射系数序列的效果,揭示了谱反演补偿频谱的过程。谱反演能够分辨出小于调谐厚度的薄层,提高了地震勘探的分辨率；谱反演具有恢复复杂地质模型的能力，适用于复杂、隐蔽油气藏的勘探。

Abstract

The basic principle of spectral inversion was elaborated and the realization workflow was given.By the application analysis of spectral inversion on single-trace model,wedge model and Mamousi model,we verified the effect of removing the impact of seismic wavelet from seismic record and obtaining reflection coefficient sequence,and revealed the compensation of frequency spectrum.Spectral inversion can identify the layer thinner than tuning thickness and improve the resolution of seismic exploration.In addition,spectral inversion can invert complex geologic model and is suitable to prospect the subtle reservoirs.

关键词：

反射系数反演; 谱反演; LSQR; 奇偶分解; 地震薄层;

Keywords:

reflection coefficient inversion; spectral inversion; Least Square QR(LSQR); odd-even decomposition; seismic thin-bed;

基金项目

DOI

10.3969/j.issn.1000-1441.2012.01.002