关键词: 四参数随机生长法, 分形维数, 自相关距离, 多元回归拟合

Abstract: The quartet-parameter structure generation set (QSGS) method is a commonly used modeling algorithm for porous media. The fractal dimension D and pore autocorrelation distance λ  of porous media are important indicators for evaluating modeling effect. There is currently a lack of research on the relationship between the parameters of the QSGS method and the fractal dimension, as well as the pore autocorrelation distance of porous media. A method for normalizing the probability distribution of the cores of growth phase is proposed, and a set of numerical simulation schemes for modeling porous media with different grid sizes N, porosities φ and normalized distribution probability p_η  is designed on this basis. Then the variations of D and λ  with the modeling parameters N, φ and p_η are analyzed based on the numerical simulation results, and the sensitivity of the modeling parameters is analyzed based on the orthogonal tests. Finally, the fitting relationships between the modeling parameters N、φ 、 p_η  and D、λ are established. The results show that the modeling parameters N、φ  and p_η all have a certain degree of influence on D, and the degree of influence is φ 、 p_η  and N in descending order. The main influencing factor of λ  is p_η , and the other parameters have less influence. The constructed multiple regression fitting models have high precision and can be used to guide the modeling of the QSGS method.

Key words: quartet-parameter structure generation set method, fractal dimension, autocorrelation distance, multiple regression fitting