摘要
近年来,乌鲁木齐市雾霾现象频繁出现.而PM 2.5浓度的高低是出现雾霾的首要因素,为了提前预测PM 2.5浓度值并及时通过人工降水法降低其浓度,对乌鲁木齐市空气质量监测站收集到6个月的观测数据进行分析,得到与乌鲁木齐市PM 2.5浓度强相关的因素,为综合考虑天气、温度、风力风向对PM 2.5浓度观测值的影响,以添加虚拟变量及因子分析提取因子的方法,将强相关因素、提取的天气、温度、风力风向因子与PM 2.5浓度观测值进行回归预测,通过模型检验得到模型的预测效果较好.
In recent years, Urumqi haze phenomena the primary factor, in order to predict the PM 2.5 density occur frequently. And PM 2.5 concentration is a haze of early and reduce its concentration by artificial precipita- tion method. Articles of Urumqi city air quality monitoring station to 6 months of observation data were collected for analysis, get the factors associated with PM 2.5 concentrations in Urumqi is strong, for considering the weather, tem- perature, wind direction on the concentrations of PM 2.5 observed value, to add a virtual variable and the method of factor analysis to extract factor, will strongly related factors, the extraction of the weather, temperature, wind direc- tion factor and PM 2.5 observed concentration value regression prediction, prediction effect of the model is obtained by model test is better.
出处
《伊犁师范学院学报(自然科学版)》
2015年第4期48-53,共6页
Journal of Yili Normal University:Natural Science Edition
基金
新疆维吾尔自治区研究生科研项目<乌鲁木齐PM 2.5浓度的动态分析研究>(XJGRI2015120)
关键词
PM
2.5
虚拟变量
因子分析
多元回归
PM 2.5
Virtual variables
Factor analysis
Muhiple regression