building N times particle gradation models and property models including two main part:
-
building lithological model read N randomly generated sets of Particle Gradation(PG) data, and interpolate them respectively. ① read each '.txt' file iteratively ② estimate Variogram parameters for each PG endmember and interpolate it ③ abstract lithological model from N sets of PG models (This part is not included in the code)
-
predicting property models read N sets of PG models and w, e models to precdict each property model ① select one property sample (training) data to be estimated and read it(in '.txt' file form), ② train the BP-ANN model ③ read w,e models and current set of PG model, using the trained BP-ANN model to predict property model ④ iteratively read N sets PG models, we can get N sets of property models run above code for properties of a1-2,a2-3,Es1-2,...,Ip, respectively, then we can obtain each property's model.