Application Research Of Hilbert-huang Transform In Logging Data Processing

Currently,most of the oil and gas field geological research is based on conventional logging data.In order to break through the limitation of the analysis of the “visual field” of the conventional logging curves,and “mining” more geological information from conventional logging data,in this paper,Hilbert-Huang transform is applied to conventional logging data processing,thus providing a new method for the interpretation of the logging data,so as to serve the study of the logging geology.Firstly,the basic principle of Hilbert-Huang transform is introduced in this paper.The decomposition process of EMD algorithm and the stopping criterion are emphasized,and the advantages and disadvantages of Hilbert-Huang transform are summarized,the simulation signal is analyzed.Secondly,the logging curve de-noising analysis based on Hilbert-Huang transform is carried out,three methods of HHT reducing denoising,wavelet threshold denoising and HHT-WT combined denoising are compared and analyzed,and change the intensity of noise to carry out research.The sequence stratigraphic division is carried out based on Hilbert-Huang transform,the IMF components obtained by the EMD decomposition can be used to divide the sequence boundaries and stratigraphic cycles.For the identification of reservoir fluid,the RILD curve is analyzed,and the combination of IMF2 and Amp2 can be used to identify the oil and water interface.Finally,the application effect of Hilbert-Huang transform in the processing of logging data are analyzed and summarized.The results show that the core part of the Hilbert-Huang transform is the empirical mode decomposition(EMD),the method is simple,highly efficient and self adaptive;When the noise intensity is weak,wavelet threshold denoising and HHT-WT combined denoising have good effect,when the noise intensity is moderate,the three denoising methods all can get better results,when the noise intensity is high,the wavelet threshold denoising is the preferred method;The IMF with high frequency can be divided into small sequence(short term cycle),the IMF with intermediate frequency can be divided into medium sequence(intermediate cycle),the IMF with low frequency can be divided into large sequence(long term cycle);At the interface of oil and water,the IMF2 component reached the peak value,and the Amp2 component reached the maximum amplitude.The use of this method should pay attention to eliminate the interference of interbed to the logging curve,and focus on the comprehensive application of a variety of logging series.