Neural Network Modeling and Prediction of Daily Average Concentrations of PM10, NO 2 and SO2

Abstract

Three-layer principal component based artificial neural network (ANN) model is used to predict PM10, NO2 and SO2 concentration. The developed model predictions are compared with the measured pollutant concentrations. The daily average pollutant concentrations and five meteorological variables are used to develop pollution forecast models. The selected monitoring site is a typical residential area with high traffic influence and the air pollution is because of nearby industries. A principal component regression (PCR) model is used for comparing the results obtained by the developed neural network model. The performance of the developed model were assessed using various performance index. Developed models exhibit a decent performance >70–95% for three measured pollutants. The future models performed with good accuracy and the predicted pollutant concentrations were confirmed to be adequate after computing the accuracy using performance indicators.