缺失地震信息重构方法研究与应用
发布时间：2018-07-20 14:10
【摘要】：随着人类社会迈入21世纪,能源的开发与应用已经关系到日常生活的方方面面。而石油和天然气,作为工业的“黄金血液”,在多种多样的能源形式中仍占有重要的地位,其对国家的经济建设、社会发展以及国防安全都有着不可估量的作用。经济的快速发展对石油和天然气的需求与日俱增,也推动了油气勘探在高新技术领域的发展,使得地球物理技术迅猛发展、地震地质理论不断推陈出新。复杂多变的地质结构和艰苦的野外勘探环境给地震勘探采集、地震资料数据处理和地震资料解释等一系列工作提出了严峻的挑战。由于受到环境及仪器等因素的限制,采集的地震资料往往缺失很多重要信息,存在采样不完整、高频成分缺失等现象,无法满足精细地震解释处理技术的要求。本文从地震数据重建和地震信号拓频这两个方面提出重缺失地震信息重构算法,从而解决地震数据采样不完整和高频成分缺失的问题,主要工作概括如下:1.对凸集投影算法(Projection onto convex sets,POCS)进行改进:传统的POCS算法具有实现原理简单,插值重建效果好的优点,但存在着计算效率低,算法收敛慢的缺陷。针对以上的不足,对传统的POCS算法的阈值策略从阈值下降模型、阈值函数、迭代停止准则三个角度提出相应的改进方案,同时结合前人的研究工作对传统的POCS算法的实现流程进行修正,精简冗余的步骤,进一步提高算法的计算效率。通过对合成数据和实际地震数据的仿真实验,验证POCS算法的有效性和普适性,并探索该方法在高维地震数据插值重建方面的应用效果。2.提出一种空间相关的非平稳地震信号拓频方法:由于受到大地滤波作用和地层吸收作用的影响,地震信号高频成分缺失严重,导致地震资料的分辨率较低。同时,地震信号作为一种短时、突变的非平稳信号,其频谱随时间变化。针对地震信号的非平稳特征,本文提出空间相关的非平稳地震信号拓频方法,该方法利用常用的时频分析工具S变换,并结合地震吸收补偿理论与空间多道统计思想,在有效地补偿高频成分,提高地震分辨率的同时,能够很好保持地震信号横向上的连续性。通过实际地震数据验证该方法能够达到预期的恢复高频成分的目的。
[Abstract]:As the human society enters the 21st century, the development and application of energy have been related to every aspect of daily life. Oil and natural gas, as the "golden blood" of industry, still play an important role in various forms of energy, and play an inestimable role in national economic construction, social development and national defense security. With the rapid development of economy, the demand for oil and gas is increasing day by day, which also promotes the development of oil and gas exploration in the field of high and new technology, and makes the geophysical technology develop rapidly. The complex and changeable geological structure and the difficult field exploration environment pose a serious challenge to seismic exploration acquisition, seismic data processing and seismic data interpretation. Due to the limitation of environment and instruments, the collected seismic data often lack a lot of important information, such as incomplete sampling and lack of high-frequency components, which can not meet the requirements of fine seismic interpretation and processing technology. In this paper, an algorithm for reconstructing seismic information is proposed from the aspects of seismic data reconstruction and seismic signal continuation, so as to solve the problems of incomplete seismic data sampling and high frequency component loss. The main work is summarized as follows: 1. The projection onto convex projection algorithm (POCs) is improved: the traditional POCs algorithm has the advantages of simple implementation principle and good effect of interpolation reconstruction, but it has the defects of low computational efficiency and slow convergence. In view of the above shortcomings, the threshold strategy of the traditional POCs algorithm is improved from three aspects: threshold descending model, threshold function and iterative stopping criterion. At the same time, combined with the previous research work, the traditional POCs algorithm is modified, the redundant steps are simplified, and the computational efficiency of the algorithm is further improved. The validity and universality of POCs algorithm are verified by the simulation of synthetic data and actual seismic data, and the application effect of this method in interpolation reconstruction of high-dimensional seismic data is explored. In this paper, a spatial correlation method for frequency extension of non-stationary seismic signals is proposed. Due to the influence of earth filtering and stratigraphic absorption, the high frequency components of seismic signals are seriously missing, which leads to the low resolution of seismic data. At the same time, the frequency spectrum of seismic signal, as a kind of short-time and abrupt non-stationary signal, varies with time. In view of the non-stationary characteristics of seismic signals, this paper presents a spatially correlated method for frequency extension of non-stationary seismic signals. The method uses S-transform, a common time-frequency analysis tool, and combines the theory of seismic absorption compensation with the idea of spatial multichannel statistics. The high frequency component can be compensated effectively and the seismic resolution can be improved at the same time the horizontal continuity of seismic signal can be maintained. The actual seismic data show that the proposed method can achieve the desired recovery of high frequency components.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4
本文编号：2133794
[Abstract]:As the human society enters the 21st century, the development and application of energy have been related to every aspect of daily life. Oil and natural gas, as the "golden blood" of industry, still play an important role in various forms of energy, and play an inestimable role in national economic construction, social development and national defense security. With the rapid development of economy, the demand for oil and gas is increasing day by day, which also promotes the development of oil and gas exploration in the field of high and new technology, and makes the geophysical technology develop rapidly. The complex and changeable geological structure and the difficult field exploration environment pose a serious challenge to seismic exploration acquisition, seismic data processing and seismic data interpretation. Due to the limitation of environment and instruments, the collected seismic data often lack a lot of important information, such as incomplete sampling and lack of high-frequency components, which can not meet the requirements of fine seismic interpretation and processing technology. In this paper, an algorithm for reconstructing seismic information is proposed from the aspects of seismic data reconstruction and seismic signal continuation, so as to solve the problems of incomplete seismic data sampling and high frequency component loss. The main work is summarized as follows: 1. The projection onto convex projection algorithm (POCs) is improved: the traditional POCs algorithm has the advantages of simple implementation principle and good effect of interpolation reconstruction, but it has the defects of low computational efficiency and slow convergence. In view of the above shortcomings, the threshold strategy of the traditional POCs algorithm is improved from three aspects: threshold descending model, threshold function and iterative stopping criterion. At the same time, combined with the previous research work, the traditional POCs algorithm is modified, the redundant steps are simplified, and the computational efficiency of the algorithm is further improved. The validity and universality of POCs algorithm are verified by the simulation of synthetic data and actual seismic data, and the application effect of this method in interpolation reconstruction of high-dimensional seismic data is explored. In this paper, a spatial correlation method for frequency extension of non-stationary seismic signals is proposed. Due to the influence of earth filtering and stratigraphic absorption, the high frequency components of seismic signals are seriously missing, which leads to the low resolution of seismic data. At the same time, the frequency spectrum of seismic signal, as a kind of short-time and abrupt non-stationary signal, varies with time. In view of the non-stationary characteristics of seismic signals, this paper presents a spatially correlated method for frequency extension of non-stationary seismic signals. The method uses S-transform, a common time-frequency analysis tool, and combines the theory of seismic absorption compensation with the idea of spatial multichannel statistics. The high frequency component can be compensated effectively and the seismic resolution can be improved at the same time the horizontal continuity of seismic signal can be maintained. The actual seismic data show that the proposed method can achieve the desired recovery of high frequency components.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4
【参考文献】
相关期刊论文 前3条
1 刘喜武;年静波;刘洪;;基于广义S变换的地震波能量衰减分析[J];勘探地球物理进展;2006年01期
2 陈学华;贺振华;;改进的S变换及在地震信号处理中的应用[J];数据采集与处理;2005年04期
3 崔树果;朱凌燕;王建花;;f-x域Cadzow技术分块压制随机噪声及其应用[J];石油物探;2012年01期
，本文编号：2133794
本文链接：https://www.wllwen.com/kejilunwen/diqiudizhi/2133794.html
