基于ARIMA模型的PM_(10)的预测与分析

The Prediction and Analysis of PM_(10) based on the ARIMA Model

通过分析北京古城2014年4月8日至7月14日连续14周污染指标PM10的监测数据,运用时间序列分析方法,借助SPSS软件对未来2周即2014年7月15—28日空气污染物变化情况进行了预测。通过Ljung-Box Q检验以及拟合统计量R2说明ARIMA模型预测效果很好,能用来对未来空气污染物的变化规律进行很好的描述,具有推广价值。

Based on the analysis of monitoring data on PM10 pollution index- es of six regions in Beijing from April 8 of 2014 to July 14 of 2014, this article adopted time series analysis method to forecast the change of air pollutants in the next two weeks （July 15 of 2014 to July 28 of 2014） with the help of SPSS. It was indicated that the predicted effect of ARIMA model was very good through the Ljung-Box Q test and fitting statistic R-square. The ARIMA model could be used to describe the change rules of air pollutants in the future, and was worth popularizing.