基于GM-ARMA组合模型的PM2.5浓度预测——以扬州市为例  

The Forecast of PM 2.5 Concentration Based on GM-ARMA Model——Taking Yangzhou as an Example

作　　者：徐辉军[1] 张林男 XU Hui-jun;ZHANG Lin-nan(Department of Basic Courses,Yangzhou Polytechnic Institute,Yangzhou 225127,China)

机构地区：[1]扬州工业职业技术学院基础部,江苏扬州225127

出　　处：《南通职业大学学报》2018年第4期67-71,共5页Journal of Nantong Vocational University

基　　金：扬州工业职业技术学院校级科研课题(2016xjrw018);扬州工业职业技术学院校级"青蓝工程"资助项目;江苏高校哲学社会科学研究基金项目(2017SJB1189)

摘　　要：以扬州市逐日空气PM2.5浓度数据为研究对象,收集2014年至2017年PM2.5月浓度时间序列,构建GM-ARMA组合模型,对其PM2.5浓度变化进行了分析和预测。研究结果表明:与GM(1,1)、AMRA(2,1)模型相比,GM-ARMA组合模型具有更好的拟合效果,对PM2.5浓度预测精度更高。Taking the daily PM 2.5 concentration of Yangzhou City as the research object,this paper collects monthly data of PM2.5 concentration from 2014 to 2017 and makes analysis and prediction of the change of PM2.5 concentration based on GM-ARMA model.The results show that compared to the GM(1,1)model and AMRA(2,1)model,with the GM-ARMA model the better fitting effect and higher accuracy of forecast of PM 2.5 concentration could be achieved.

关 键 词：PM2.5 时间序列 GM-ARMA组合模型 预测 

分 类 号：O29[理学—应用数学] N945[理学—数学]