先验信息构建与地震随机反演方法研究

发布时间：2018-05-22 11:27

本文选题：先验信息构建 + 地震随机反演　；参考：《中国石油大学(华东)》2015年硕士论文

【摘要】：地震随机反演是一种结合先验信息构建与优化理论的非线性反演方法。与确定性反演相比,基于统计先验信息的地震随机反演方法可以有效地融合测井数据中包含的高频信息,具有更高的垂向分辨率。但是,常规的地震随机反演存在运行速度慢和横向连续性差的缺陷。本文研究了快速的先验信息构建方法——基于傅里叶滑动平均(FFT-MA)谱模拟和逐步变形算法(GDM)的地质统计先验信息构建与基于分形理论和克里金插值的混合先验信息构建,以及改进的扰动方式和优化策略,从而提高地震随机反演的计算速度;并且提出与确定性反演方法相结合,一定程度上提高了随机反演结果的横向连续性趋势。在获得高品质地震和测井数据的前提下,基于贝叶斯理论,地震随机反演通过将地质统计学知识、分形理论和常规的确定性反演技术相结合,充分地融合了已知的地震资料、地质信息和测井数据,使随机反演结果更加符合地质结构的框架约束。通过随机模拟方法构建反映空间不确定性特征的储层参数模型,然后利用地震反演和非线性寻优算法对随机模拟结果进行筛选,为地震解释提供了可靠的储层参数信息,具有非常重要的实际意义。本文提出的随机反演方法均直接采用精确的Zoeppritz方程计算合成地震记录,减少了方程近似所引起的误差,产生的反演结果更准确。油气藏勘探和储层评价的重要任务是进行储层预测和流体识别。本文研究了如何从有限的地震资料、测井数据及地质背景中提取储层弹性参数、物性参数和流体敏感因子,可以更全面地进行油气藏的勘探与开发。通过岩石物理理论获得岩石骨架和孔隙流体性质的定性评价,从而得到地震资料与物性参数、流体因子之间的关系,进而实现了储层物性参数和流体因子的随机反演,并且基于地震随机反演实现储层参数的直接同步估算,减少了不确定性的传递和误差的累积。本文分别利用模型和实际资料对提出的地震随机反演方法进行了测试,结果表明,这些反演方法不仅具有理论意义,而且具有一定的实用价值。
[Abstract]:Seismic random inversion is a nonlinear inversion method combined with prior information construction and optimization theory. Compared with deterministic inversion, stochastic seismic inversion based on statistical prior information can effectively fuse the high frequency information contained in logging data and have higher vertical resolution. However, the conventional random seismic inversion has the defects of slow running speed and poor horizontal continuity. In this paper, a fast method of constructing prior information is studied, which is based on Fourier moving average (FFT-MA) spectral simulation and stepwise deformation algorithm (GDM), and mixed priori information based on fractal theory and Kriging interpolation. The improved perturbation method and optimization strategy can improve the computing speed of seismic random inversion, and combine with deterministic inversion method to improve the horizontal continuity trend of random inversion results to a certain extent. On the premise of obtaining high quality seismic and logging data, based on Bayesian theory, seismic random inversion combines geostatistical knowledge, fractal theory and conventional deterministic inversion technology to fully integrate known seismic data. Geological information and logging data make the random inversion results more consistent with the framework constraints of geological structure. The reservoir parameter model reflecting spatial uncertainty is constructed by stochastic simulation method, and then the random simulation results are screened by seismic inversion and nonlinear optimization algorithm, which provides reliable reservoir parameter information for seismic interpretation. Has very important practical significance. The stochastic inversion methods proposed in this paper directly use accurate Zoeppritz equation to calculate synthetic seismic records, which reduces the error caused by the equation approximation, and the inversion results are more accurate. The important task of reservoir exploration and reservoir evaluation is reservoir prediction and fluid identification. This paper studies how to extract reservoir elastic parameters, physical property parameters and fluid sensitivity factors from limited seismic data, logging data and geological background, so that the exploration and development of oil and gas reservoirs can be carried out more comprehensively. Through the theory of rock physics, the qualitative evaluation of rock skeleton and pore fluid properties is obtained, and the relationship between seismic data and physical parameters and fluid factors is obtained, and the stochastic inversion of reservoir physical parameters and fluid factors is realized. And the direct synchronous estimation of reservoir parameters is realized based on seismic random inversion, which reduces the transfer of uncertainty and the accumulation of errors. In this paper, the seismic stochastic inversion methods are tested by using the model and the actual data respectively. The results show that these inversion methods are not only of theoretical significance, but also of practical value.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4

【相似文献】