摘要
钻井过程中钻头震动井壁会形成较浅的孔洞与划痕,在电成像静态图中表现为麻点噪声或颜色相对较深的条、块状干扰,影响裂缝的识别和参数的提取。为了降低背景噪声的影响,将联合经验模态分解法(EMD)及小波阈值去噪方法应用到电成像测井的电导率曲线中。结果表明:背景噪声主要存在于电导率曲线的高频部分,把电导率数据EMD分解后,对得到的高频固有模态函数(IMF)分量进行小波阈值去噪可有效减少背景噪声,输出的FMI(地层微电阻率扫描成像)静态图中麻点噪声和条、块状干扰明显减少;由去噪电成像静态图计算得到的缝洞面孔率与岩心孔隙度有更好的线性关系,与常规资料孔隙度和人工拾取的裂缝面孔率具有一致性,说明基于EMD的小波阈值去噪方法在电成像测井数据中的应用是有效的。
Bits quaked in the course of drilling result in shallower holes and scratches in the surfaces of borehole walls.This usually leads to noise pits or the interference of strips and blocks with darker colors that will affect the identification of fractures and extraction of parameters.In order to reduce this specific background noise,a de-noising method combining the empirical mode decomposition(EMD)and wavelet threshold processing is applied to the conductivity curve of the electrical imaging logging.The analysis show that the background noise mainly exists in the high frequency signal of conductivity curve.The electrical conductivity data of electrical imaging logging is subjected to the EMD,and the intrinsic mode function(IMF)components obtained are subjected to wavelet threshold de-noising,which can significantly reduce the background noise.There is a better linear relationship between the fracture-vug plane porosity extracted from electrical imaging static image and core porosity,and fracture-vug plane porosity is consistent with the porosity of conventional logging data and fracture plane porosity picked up artificially.The results show that the wavelet threshold de-noising method based on EMD is effective in the application of electrical imaging logging data.
作者
徐方慧
王祝文
刘菁华
欧伟明
XU Fanghui;WANG Zhuwen;LIU Jinghua;OU Weiming(College of GeoExploration Science and Technology,Jilin University,Changchun 130026,China)
出处
《中国石油大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2020年第3期56-65,共10页
Journal of China University of Petroleum(Edition of Natural Science)
基金
国家自然科学基金项目(41874135)。
关键词
裂缝性火成岩储层
电成像测井数据
背景噪声
经验模态分解法(EMD)
小波阈值去噪
缝洞面孔率
fractured volcanic formation
electrical imaging logging data
background noise
empirical mode decomposition(EMD)
wavelet threshold de-noising
fracture-vug plane porosity
