Estimating phase behavior of the asphaltene precipitation by GA-ANFIS approach

This study implements an adaptive neuro-fuzzy inference system (ANFIS) approach to predict the precipitation amount of the asphaltene using temperature (T), dilution ratio (R-v), and molecular weight of different n-alkanes. Results are then evaluated using graphical and statistical error analysis methods, confirming the model's great ability for appropriate prediction of the precipitation amount. Mean squared error and determination coefficient (R-2) values of 0.036 and 0.995, respectively are obtained for the proposed ANFIS model. Results are then compared to those from previously reported correlations revealing the better performance of the proposed model.