Short-term mathematical prediction model of air quality based on CEEMD-ELM-PSO

Abstract

After entering the new era, people's living standard has been significantly improved, the concept of environmental protection has been deeply rooted. People pursue a greener and healthier lifestyle, and the concern for air quality has become more and more intense. People need to continuously analyze relevant mechanisms to help people predict air environment quality effectively and prevent relevant hazards in time. It is difficult to obtain a better prediction effect under different AQI fluctuation trends in the traditional single machine learning model for air quality prediction. In order to solve the problem, the prediction method is improved, and the ELM-PSO algorithm is used to predict the future AQI, which helps to analyze the future air change trends from a macro perspective. This bit combines Beijing's short-term air quality data and proposes an air quality short-term prediction model based on complementary ensemble empirical modal decomposition and an optimal limit learning machine. Firstly, the data are decomposed by CEEMD to reduce the non-smoothness of the data, thus reducing the impact of non-smoothness on the prediction accuracy; then ELM model prediction is performed for each decomposed series obtained; the output weights of the limit learning machine are used to PSO optimization search, and then construct the prediction model based on the limit learning machine; finally, all the prediction components are superimposed to obtain the final results. The results surface that the prediction model proposed in this paper has high accuracy in short-term air quality prediction.