Application of Adaptive Neuro-Fuzzy Inference System Model for Predicting CO2 Emission Based on Production Rate in Sugarcane Agroindustry

Abstract

Carbon dioxide (CO₂) emissions from industrial processes, particularly in sugar production, have become an important environmental concern that requires accurate prediction and monitoring. This study aims to predict carbon dioxide (CO₂) emissions using the Adaptive Neuro-Fuzzy Inference System (ANFIS) model based on production rate and the fuel used for operating the boiler machine. The model was developed using hypothetical data obtained from the literature, which were divided into 80% training data and 20% testing data. A trial-and-error approach was employed to determine the optimal parameters of the ANFIS model. Various membership function (MF) types and numbers were evaluated, and the optimal configuration was found to be three MFs with a triangular MF type. Model performance was assessed using Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and the coefficient of determination (R²). The results show that the ANFIS model provides good predictive performance, with an RMSE value of 17.083, a MAPE of 14.338, and an R² of 0.8686. These findings indicate that the proposed ANFIS model has a high capability in estimating CO₂ emissions based on production rates and fuel usage.