Modeling and optimization of NO emission for a steam power plant by data‐driven methods

Abstract

The evolution of the power industry toward large-scale automation and selfmonitoring provides the opportunity to optimize the technical and environmental performance of the plant with data-driven methods with little changes in infrastructure. This article applies the artificial neural network (ANN) and genetic algorithm (GA) to predicting and optimizing NO emissions. Multiple linear regression models, correlation matrix, and research background are employed to find the most influential input features. The generated power, natural gas flow, the flow of gas recirculation fan, gas air heater temperature, and the amounts of oxygen in the stack are identified as the effective input features. Mean Square Error (MSE) and the coefficient of determination (R 2 ) of best architecture (22 neurons in a hidden layer) are calculated 0.0117 and 0.96651, respectively. The Average Percentage Error (APE) is usually below 10%, meaning the model is in good agreement with real data. Finally, the Genetic Algorithm (GA) is used to minimize the amount of NO emissions. The ANN-GA techniques reduce the NO emission by at least 32% for selected records, enabling us to optimally find the prominent features affecting NO emission by the operational conditions and low economic costs.

Data-Driven Modeling for the Prediction of Stack Gas Concentration in a Coal-Fired Power Plant in Türkiye

“…Similarly, MLP was not a very good fit for dust. The R 2 values in the CO and dust data to which the MLP model was applied are far from 1.0, which means that the values predicted by the model with these data and the actual values were quite far from each other (Movahed et al, 2023;Tang et al, 2022). As it can be seen from the R 2 values, the predicted and the actual data of O 2 , NOx, and SO 2 levels were closer to each other when using MLP model.…”

“…Due to the capacity of AI to detect complex temporal patterns and non-linear correlations in data, these models are more accurate in predicting the stack-gas emissions of coal-fired power plants than conventional techniques (Alnaim et al, 2022;Laubscher, 2019). Previous studies, such as those by Krzywanski & Nowak, 2016, Vujić et al, 2019, Tang et al, 2022, Chikobvu & Mamba, 2023, Movahed et al, 2023, and Josimović et al, 2023 have explored AI-driven models for forecasting specific pollutants such as CO, NOx, NO 2 , PM 2.5 , and SO 2 . These researches have primarily focused on individual pollutants, resulting in a limited number of studies that have comprehensively addressed the forecasting of all pollutants emitted from coal-fired power plants.…”