PM2.5 concentration prediction based on EEMD-Stacking - A case study of Yangtze River Delta, China

Abstract

With the acceleration of China's industrialization process, the resulting environmental problems have become increasingly prominent, especially the rising concentration of PM2.5 in the air, which has caused various consequences for people's clothing, food, housing and transportation. Due to the randomness and complexity of PM2.5 concentration time series, this paper uses EEMD to decompose the historical PM2.5 concentration data into EIMF and trend series. Considering air quality factors and meteorological factors, this paper constructs EEMD-Stacking model, and uses Bayesian algorithm to optimize the parameters. The Yangtze River Delta region was selected as the experimental site, and the daily PM2.5 concentration data and meteorological station data from 2018 to 2020 were used for prediction experiments. The results show that the combined model has good prediction effect. The short-term prediction accuracy is relatively high, and the medium and long-term prediction accuracy decreases, but the overall prediction accuracy is high and stable.