Study On The Fusion Technology Of Geophysical Attributes

Aiming at the reservoir characteristics of SL work area and improving the accuracy of lithology prediction,a series of researches are carried out around scale fusion,geologic fusion and parameter fusion.Based on the mismatch factors of logging and seismic data which affect reservoir prediction precision,a fine layer calibration method is proposed,which realizes high precision and deep time conversion.This paper analyzes and compares several methods of attribute optimization,and selects the Lasso regression supervised algorithm to optimize the seismic attributes.The support vector machine regression algorithm is introduced to solve the problem of less sample size caused by drilling.In order to overcome the problem of regression parameter optimization of support vector machine,It is proved that the method combined with genetic algorithm can effectively improve the accuracy of lithology prediction of wells through the model calculation and comparison with other methods.In order to obtain the comprehensive information of two different lithology parameters of the reservoirs,the parameter fusion rules based on the nonsubsampled contourlet transform are studied,and the useful information of reservoir lithology parameters is obtained.