关键词: 分维, 孔隙结构, 毛管压力, 相对渗透率, 模糊识别, BP神经网络

Abstract: The expression of the capillary pressure and relative permeability can be used to calculate the fractal dimension of the reservoir rock pore structure. geological characters of Reservoir are revealed by logging data. Based on this information and the method of fuzzy identification, fuzzy belongingfunctions and equations are derived for the low porosity and low permeability reservoirs of Silurian in the block. The types of reservoirs can thus be determined according to the belonging degree. On the basis of investigating reservoir′s characteristics of lithologic, physical, electrical and oilbearing properties and the relationship of them, a logging classification model of identifying reservoir type is established by use of BP neural network method, and the prediction accuracy is tested. The assessment parameters are analyzed from 10 wells in target zone with the methods, the result is much better than ever before. The research indicates fractal dimensions of the pore structure can quantitatively characterize the reservoir property of sandstones. The types of the pore structure with different origin have different fractal dimensions, so fractal dimension is of importance in classification and evaluation of the pore structure in sandstones.

Key words: fractal dimension, pore structure, capillary pressure, relative permeability, fuzzy identification, BP neural network