Modeling of air quality prediction for PM2.5 concentration in Chengdu area based on measured data

Abstract

Air quality, especially the concentration of PM2.5, has attracted widespread attention as it affects people's health and may cause health conditions such as allergies, coughs and lung diseases. Chengdu is located in the Sichuan Basin. The unique geographical environment and climatic conditions make Chengdu's PM2.5 changes have its own characteristics. Based on the measured data, this paper proposed a prediction model for the changes of PM2.5 in Chengdu. Specifically, using the correlation analysis between air composition and climate factors, a novel predictive model structure was constructed. Then, based on the historical measured data of air quality in Chengdu area, the parameters of the prediction model were identified using optimization algorithms. Finally, the comparison between the predicted value given by the established prediction model and the measured value of PM2.5 verified the effectiveness of the prediction model.