Contents I. Introduction
Air pollution has been a hot issue in recent years, the rapid development stage neural network provides a powerful tool for the prediction of atmospheric pollutants. Artificial neural network simulated by the way the brain processes the problem, the information processing and nonlinear conversion, and an adaptive feedback error correction. Because of its unique ability and some parallel processing of information, self-learning and reasoning skills, in recent years it has been widely applied to various fields. This paper fully consider the factors that affect PM2.5 levels, such as SO2, NO2, etc., genetic algorithm optimization BP neural network model depicts the occurrence and evolution of PM2.5 in Changsha, the training data 2016.01-2018.04, and generating prediction data, the actual data is relatively 2018.05-2018.06 than satisfactory.
