CEEMD-SG滤波在固井质量测井资料处理中的应用

Application of the CEEMD-SG fltering method in well cementing quality logging data processing

目前固井质量评价主要依靠声波变密度测井资料,受双层套管影响,变密度图像有时会因内、外层套管波叠加出现明显的畸变现象,对固井质量的评价产生干扰。为获得真实的变密度图像,剔除“假异常”,将互补集合经验模态分解(CEEMD)和Savitzky-Golay(SG)滤波法联合,提出一种CEEMD-SG滤波方法,并应用于声波变密度测井数据处理。首先使用CEEMD方法对变密度数据进行分解,通过傅里叶变换对原始信号及分解得到的各个固有模态函数(IMF)进行频谱分析,然后选取高频IMF分量使用SG滤波算法进行处理,并将滤波后的高频分量与CEEMD分解得到的低频分量联合重构得到处理后的信号。利用该方法对塔里木油田实际变密度资料进行处理,并与CEEMD、小波阈值去噪、SG滤波三种方法进行对比,结果表明,CEEMD-SG滤波方法能更好地消除双层套管带来的波形叠加干扰,处理后的资料能满足现场验收要求。

At present,the evaluation of cementing quality mainly relies on acoustic variable density logging data.Due to the influence of double-layer casing,the variable density image may sometimes exhibit significant distortion due to the superposition of inner and outer casing waves,which can interfere with the evaluation of cementing quality.In order to obtain real variable density images and eliminate"false anomalies",a complementary set empirical mode decomposition(CEEMD)and Savitzky-Golay(SG)filtering method are combined to propose a CEEMD-SG filtering method for processing acoustic variable density logging data.Firstly,the CEEMD is used to decompose the variable density data,and then the Fourier transform is used to analyze the original signal and each intrinsic mode function(IMF)obtained from the decomposition.High frequency IMF components are selected for processing using the SG filtering algorithm.Finally,the filtered high frequency components and the low frequency components obtained from CEEMD decomposition are combined to reconstruct the processed signal.Using this method to process the actual variable density data of Tarim Oilfield,and comparing it with three methods:CEEMD,wavelet threshold,and SG filtering,the results show that the CEEMD-SG filtering method can better eliminate the waveform superposition interference caused by double-layer casing,and the processed data can meet the on-site acceptance requirements.