Application of adaptive neuro-fuzzy inference system for prediction of dissolved oxygen concentration in the gold cyanide leaching process

An adaptive neuro-fuzzy inference system (ANFIS) model has been developed for the prediction of the dissolved oxygen concentration (DOC) as a function of the solution temperature (0-40°C), salinity based on conductivity (0-59000 μS/cm), and atmospheric pressure (600-795 mmHg). The data set was randomly divided into two parts, training and testing sets. 80% of the data points (80% = 11556 datasets) were utilized for training the model and the remainder data points (20% =2889 datasets) were utilized for its testing. Several indices of performance such as root mean squared error (RMSE), mean absolute percentage error (MAPE), and coefficient of correlation (R) were used for checking the accuracy of data modeling. ANFIS models for the prediction of DOC were constructed with various types of membership functions (MFs). The model with the generalized bell MF had the best performance among all of the given models. The results indicate that ANFIS is a powerful tool for the accurate prediction of DOC in gold cyanidation tanks.