淮安市PM2.5的分布特征与灰色预测模型

The Distribution Characteristics and Gray Prediction Model of PM2.5 in Huaian

研究对象是淮安市主要大气污染物中PM2．5在2013—2016年间的观测数据，从两个方面对PM2．5的浓度进行了研究．一方面，利用非参数假设检验中的卡方拟合检验与w检验等方法研究了PM2．5浓度的分布特征．研究发现，污染物PM2．5的日均浓度服从对数正态分布、月均浓度服从正态分布．另一方面，建立了PM2．5年均浓度的精度为一级可用于长期预测的GM（1，1）预测模型，利用预测模型对未来五年中PM2．5的浓度水平进行了预测．

This paper studies the observed data from 2013 to 2016 in Huaian of its PM2.5 in atmospheric pollutant and, from two aspects, the thickness of PM2.5. On the one hand, via the chi square test and W test of nonparametric test methods, this paper studies the distribution characteristics of PM2.5 concentration. The study founds that the daily average concentration of pollutants PM2.5 obeys logarithmic normal distribution, while the monthly average concentration obeys normal distribution. On the other hand, a GM（1,1） prediction of the annual average concentration of PM2.5 with the accuracy of level i has been established, which can be used for long-term prediction. The prediction of the concentration level of PM2.5 in the next five years were predicted.