地球物理学进展 ›› 2018, Vol. 33 ›› Issue (6): 2403-2408.doi: 10.6038/pg2018BB0519

• 应用地球物理学Ⅰ(油气及金属矿产地球物理勘探) • 上一篇    下一篇

基于低频软约束-基追踪薄层反演方法研究

李治昊,张丰麒()   

  1. 中国石油化工股份有限公司石油勘探开发研究院,北京 100083
  • 收稿日期:2018-03-21 修回日期:2018-09-21 出版日期:2018-12-20 发布日期:2019-03-03
  • 通讯作者: 张丰麒 E-mail:chaseramd@aliyun.com
  • 作者简介:李治昊,男,汉族,1985年生,硕士,工程师,主要从事于测井、地质以及地震等专业软件的研发研究.(E-mail:lizh. syky@sinopec.com)
  • 基金资助:
    国家科技重大专项“中国陆相页岩油资源潜力与地质评价研究”(2017ZX05049-001-003);中科院A类战略性先导科技专项(XDA14010405)

Research of a thin-bed inversion based on basic pursuit with the low-frequency soft constraint

LI Zhi-hao,ZHANG Feng-qi()   

  1. Exploration and Production Research Institute, SINPEC, Beijing 100083,China
  • Received:2018-03-21 Revised:2018-09-21 Online:2018-12-20 Published:2019-03-03
  • Contact: Feng-qi ZHANG E-mail:chaseramd@aliyun.com

摘要:

提高地震的垂向分辨率仍然是当前地震勘探阶段的重点和难点.本文通过引入反射系数奇偶分解理论,并借助于反射系数解耦矩阵将反射系数分解为奇分量和偶分量,其中偶分量对提高地震垂向分辨率具有帮助作用.在此基础上,利用信号稀疏表示领域中的基追踪技术,在最小化模型参数L1范数的约束下,可以有效求解大型欠定方程组,进而得到高分辨率的反射系数.考虑到单一的反演约束无法恢复地震缺失的低频信息,并且反演剖面平滑度相对较差,因此通过在基追踪反演目标函数中引入低频软约束项,强迫反演结果的低频成分逼近模型的低频成分,进而可以补充0~10 Hz的低频信息,使得反演剖面的垂向分辨率和横向连续性同时增强.

关键词: 垂向分辨率, 反射系数奇偶分解, 基追踪, 低频软约束项

Abstract:

It is always an important and difficult issue that how to improve the vertical resolution of the seismic data. In the framework that the reflectivity pair can decompose into an odd component and an even one, this thesis introduces a decomposition matrix to realize this process. And the even component can help to improve the vertical resolution of the seismic data. The basic pursuit, a novel method of signal sparse representation, can effectively resolve the large underdetermined equations with the constraint of minimization of the L1 norm of the model parameters. In consideration that the single constraint couldn’t stable the inversion, the low-frequency soft constraint is added into the objective function of the basic pursuit inversion, which makes the low frequency of the inversion equal to the one of the well log. And then the inverted section can own the high resolution in vertical direction and the smoothness in horizontal direction.

Key words: the vertical resolution, the odd-even decomposition of the reflectivity pair, basic pursuit, the low-frequency soft constraint

中图分类号: