基于灰色关联度和主成分分析的不同地区API综合评价

Comprehensive evaluation of API in different areas based on gray correlation and principal component analysis

为研究API与各气象要素之间的内在联系,建立API的数学模型,选取合适的气象因子对空气污染指数进行预测提供理论支撑,本文利用统计学方法分析了6个污染物指数的变化规律及其随不同气象要素的变化特征;运用灰色关联度分析法对13个气象因子与空气污染指数的相关性进行了排序;采用主成分分析方法对12个气象因子进行了降维,找出几个综合因子反映原来变量的信息,得出不同气象因子对不同污染物指标的影响程度以及影响不同地区API的主成分具有差异性的结论。对于深入认识不同地区的空气环境质量与影响因素之间的相关性,准确把握API影响因子信息进而制定有针对性的空气环境治理措施,指导API的预测具有一定的现实意义。

In order to study on the internal relations between API and the meteorological factors, establishes the API model,and selects the appropriate meteorological factors to predict and provide theoretical support to the air pollution index,this paper uses statistical methods to analyze the 6 pollutant index variation and meteorological characteristics with different elements; using grey relational analysis to sort the correlation of 13 meteorological factors and air pollution index method; use principal component analysis method of 13 meteorological factors by reducing the dimension; find out several comprehensive factors to reflect the original variable information. It is concluded that the influence of different meteorological factors on the indicators of different pollutants and the main components affecting API in different regions have differences. It is of practical significance to understand the correlation between air quality and influencing factors in different regions,accurately grasp the information of API impact factors,formulate targeted air environment control measures and then guide API prediction.