Analysis And Prediction On Hohhot Air Pollutant SO₂ And NO₂

The concentration of pollutants in the air is an important environmental quality indicator, and also closely related to the health of the residents. In recent years, the air quality in Hohhot has showed a declining trend, the pollution caused by the industrial and automobile exhaust is becoming more and more intensified, and the control and prevention of air pollution is becoming more and more important. In order to better control and prevent air pollution, it is necessary to accurately predict the air pollution. In this paper, we use multivariate linear regression model and neural network model to predict the concentration of SO2 and NO2 in the air in the future. We collected data such as emission data of 26 major factories in Hohhot, pollutant concentration data of eight air automonitoring station and daily meteorological data of Hohhot Baita airport ground station. Through analysis and data mining, we extracted the features used for modeling and constructed the training and test samples. Based on learning of the training samples and testing on the test samples, we determined the effective model parameters. After a lot of experiments, the results show that the neural network model used for one-day-ahead prediction of sulfur dioxide and nitrogen dioxide concentration in the air generates higher accuracy, can better fit the actual pollutant concentration change trend.