摘要
目的借助于自回归积分滑动平均(Autoregressive Integrated Moving Average Model,ARIMA)模型对苏州市空气质量指数进行预测,为空气污染的健康防护预警提供参考。方法运用R软件以苏州市2018年1月1日至12月31日的日空气质量指数为基础,借助于参数估计等筛选最佳的ARIMA模型,以此为基础对苏州市2019年1月1日至1月6日的空气质量指数进行预测,评价其预测效果。结果借助于2018年苏州市日空气质量指数构建了ARIMA(1,1,1)模型,模型的AIC=267.06,Box-Ljung检验的Q统计量为18.558,P=0.775,残差序列为白噪声。空气质量指数的预测值与实际值较为接近,绝对误差的平均值为-7,相对误差的平均值为-4.29%,模型预测效果比较理想。结论 ARIMA(1,1,1)模型能够较为理想的对苏州市空气质量指数进行预测,在空气质量指数预测中具有良好的应用前景。
Objective To predict the air quality index (AQI) in Suzhou with the application of Autoregressive Integrated Moving Average Model (ARIMA), and provide reference for health protection warnings for air pollution. Methods Based on the daily air quality index of Suzhou City from January 1 to December 31, using R software, the best ARIMA model was selected by means of parameter estimation. On the basis of this, the AQI of Suzhou from January 1 to January 6, 2019 was predicted to evaluate its forecasting effect. Results The ARIMA(1,l,1) model was established based on the daily AQI of 2018, and the AIC of this model was 267.06. Ljung-Box test showed that Q value was 18.558 (P=0.775), and the residual sequence was white noise. The prediction results of the AQI accorded well with the actual data. The average absolute error was -7, and the mean relative error was -4.29%. Conclusion The ARIMA (1, l, 1) model ideally predicted the AQI in Suzhou, and it has a good application prospect in the prediction of AQI.
作者
王建书
王瑛
赵敏娴
周晓龙
陆颂文
杨海兵
刘强
WANG Jianshu;WANG Ying;ZHAO Minxian;ZHOU Xiaolong;LU Songwen;YANG Haibing;LIU Qiang(Suzhou Center for Disease Control and Prevention,Suzhou,Jiangsu 215004,China)
出处
《公共卫生与预防医学》
2019年第2期18-20,共3页
Journal of Public Health and Preventive Medicine
基金
苏州市"科教兴卫"青年科技项目(KJXW2017053)
