基于ARIMA模型的杭州市PM2.5预测

Forecast of PM2.5 in Hangzhou Based on ARIMA Model

选取2017年1月1日至2017年12月31日PM2.5的日均浓度的时间序列作为样本数据,利用Box-Jenkins的建模理论,建立符合杭州市的PM2.5浓度变化的ARIMA(4,1,2)模型.利用2018年1月1日~2018年1月5日的PM2.5浓度的预测值与实际值相比较,结果表明ARIMA(4,1,2)模型预测效果较好,适用于杭州市PM2.5的预测.

Selected the time series of daily average concentration of PM2. 5 from January 1,2017 to December 31,2017 as sample data,and used the Box-Jenkins modelings theory,an ARIMA（4,1,2）model is established,which conforms to the change of PM2. 5 concentration in Hangzhou. Used the predicted values of PM2. 5 concentrations from January 1,2018 to January 5,2018,compared with the actual values,the results show that the ARIMA（4,1,2） model has a good prediction effect which is applicable to Hangzhou PM2. 5 predictions.