Numerical simulation of regional air pollution characteristics based on meteorological factors and improved Elman neural network algorithm

The traditional regional air pollution characteristics analysis method cannot obtain the best meteorological pollution factors, leading to air pollution characteristics analysis results and actual deviation. Therefore, a numerical simulation method of regional air pollution characteristics based on weather factors and improved Elman neural network algorithm is proposed. The best meteorological factors affecting air quality were obtained and the annual mean concentration of air pollutants was analyzed by Spearman rank correlation coefficient method. According to the structure of Elman neural network, an interval prediction model of atmospheric pollution concentration is designed. Based on this, control equations, turbulence model and boundary conditions are constructed. The characteristics of air pollution in a city are simulated to obtain the maximum contribution of air pollution. The results show that the simulated concentration of PM2.5and PM10 is close to the measured values, the relative error between the simulated values and the measured values is small, and the simulated atmospheric concentration is in good agreement with the measured values. Therefore, the numerical simulation analysis of pollution characteristics presented in this paper is more accurate and reliable.