秦皇岛市空气质量特征分析及其预测模型  

Characteristics Analysis and Prediction Model of Air Quality in Qinhuangdao City

作　　者：董丽沙[1] 范金圆 吕金凤[1] 孔亮[1] 张步英[1] 杨晓静[1] 李洋洋 DONG Lisha;FAN Jinyuan;Lü Jinfeng;KONG Liang;ZHANG Buying;YANG Xiaojing;LI Yangyang(School of Mathematics and Information Science&Technology,Hebei Normal Uniiversity of Science&Technology,Qinhuangdao Hebei,066004;MCC Shenkan Qinhuangdao Engineering Design and Research Institute Co.Ltd,China)

机构地区：[1]河北科技师范学院数学与信息科技学院,河北秦皇岛066004 [2]中冶沈勘秦皇岛工程设计研究院总院有限公司

出　　处：《河北科技师范学院学报》2020年第4期40-47,共8页Journal of Hebei Normal University of Science & Technology

摘　　要：以秦皇岛市为研究对象,对其2015~2019年空气质量的变化特征进行分析,得出秦皇岛市空气质量的特征和变化趋势,并建立了秦皇岛市空气质量指数的预测模型。分别采用2种预测方案,并在每种方案中选用不同的预测模型:方案1,基于AQI时间序列的乘积季节模型;方案2,基于气象因子的BP神经网络AQI预测模型。通过对比不同模型的预测及结果,得到2个模型中最适合秦皇岛市空气质量指数的预测模型为乘积季节模型。Taking Qinhuangdao City as the research object,the change characteristics of air quality from 2015 to 2019 were analyzed,and the distribution characteristics and change trend of air quality in Qinhuangdao City were obtained.Two kinds of forecasting schemes are adopted to establish the air quality prediction model of Qinhuangdao city with different prediction model each:the product season model based on AQI time series;the AQI prediction model of BP neural network based on meteorological factors.Through the comparison of the prediction and results of different models,it is concluded that the product seasonal model is the most suitable model for air quality prediction in Qinhuangdao city.

关 键 词：空气质量 变化趋势 乘积季节模型 BP神经网络模型 秦皇岛市 

分 类 号：X821.222.23[环境科学与工程—环境工程]