Prediction of Nox Emissions from Gas Turbines of a Combined Cycle Power Plant Using an Anfis Model Optimized by Ga

Combined cycle power plants, which combine gas and steam turbines, have an adverse impact on surrounding populations and structures. Input data can be fine-tuned to reduce this effect. If the parameters that affect the NOx (NOx = NO2 + NO) value measured as an output are properly determined, the system can develop in the desired direction. In this study, NOx emission was estimated by adaptive neuro-fuzzy inference system (ANFIS) using Predictive Emission Monitoring System (PEMS) data from a natural gas fired combined cycle power plant. Then, parameter optimization was performed using a genetic algorithm (GA) to reduce the prediction error. The proposed ANFIS-GA framework has been developed, trained and tested on these datasets. Various performance measures (mean square error (MSE), error standard deviation (ED), correlation coefficient (R), error mean (EM), and root mean square error (RMSE)) are used to show the capacity of the model and evaluate its performance. The results show that GA has a considerable impact on the performance of ANFIS training and significantly increases the predictive accuracy of the model. Since the model ANFIS-GA is a realistic optimization method for ANFIS models, the results obtained with this model are more accurate than the results obtained with other approaches used