摘要
基于高阶统计量具有检测和表征信号中的非线性等特点,文中提出了用高阶统计量识别和预测生物礁储集层的新方法,其目的是通过高阶统计量检测出生物礁的非高斯性信息,预测生物礁储层的存在。文中针对高阶矩对非高斯信号比较敏感,特别是偏离高斯性的程度越大、高阶矩越大的特性,将二阶相关的相干体算法类推到三阶、四阶,提取主要反映高斯偏离程度的高阶相干切片。结果表明,约90%的地震道都服从广义高斯分布。高阶相干体能较好地预测生物礁体的分布。
As the high-order statistics have the ability to detect and characterize the nonlinearity and other parameters,a new method in which the high-order statistics were used to identify and predict the reef reservoir was proposed in this paper. The propose of the research is to utilize the high-order statistics to detect the non-Gaussian property of the reef so as to predict the existence of the reef reservoir. Regarding the sensitiveness of the high-order moment to non-Gaussian signal,especially the characteristic in which the greater extent the signal deviate from the Gaussianity,the greater the high-order moment is,the second-order related coherence cube algorithm was analogized to third-order and fourth order,so the high-order coherence slices which reflect the Gussian deviation extent was extracted. The research results show that 90 percent seismic traces are subject to generalized Gaussian distribution. The high-order coherence cube could accurately predict the distribution of the reef.
出处
《石油地球物理勘探》
EI
CSCD
北大核心
2010年第5期705-709,共5页
Oil Geophysical Prospecting
关键词
生物礁
储层预测
广义高斯分布
高阶统计量
相干切片
reef,reservoir prediction,Generalized Gaussian Distribution,higher-order statistics,coherence slice
