关键词: 狼优化算法, BP神经网络, PM2.5浓度预测, 预测模型, 空气污染物

Abstract: Abstract: Focus on the issue of high cost and complicated process of the fine particulate matter (PM2.5) measurement system,the neural network model based on the grey wolf optimizer algorithm was established .From the perspective of non - mechanism model,the daily PM2.5 concentration in Shanghai was forecasted. The meteorological factors and air pollutants was considered in this paper, and the important factors were analyzed by mean influence value. To avoid the "over training" and ensure the generalization ability ,the validation datasets were used to monitor the training process. The experimental results show that the most significant factors affecting the PM2.5 concentration was PM10 ,and the CO and the previous day PM2.5 was followed .Based on the datasets obtained from November 01, 2016 to November 12, 2016, the relative average error decreases by about 5 percentage points compared with BP neural network prediction model, and decreases by about 3 percentage points compared with Particle Swarm Optimization(PSO). The neural network model based on the grey wolf optimizer algorithm is more suitable for forecasting PM2.5 concentration and air quality in Shanghai.

Key words: grey wolf optimizer algorithm, BP neural network, PM2.5 concentration prediction, prediction model, air pollutants