基于SOM网络及关联规则的湖北省城市空气质量研究

Urban Air Quality Evaluation of Hubei Province Based on SOM Network and Association Rules

摘要: 随着经济发展及人口增多，我们居住的城市陆续发生了各种环境问题。利用自组织竞争(SOM)神经网络研究湖北省城市环境空气质量，以PM2.5、PM10、SO2、CO、NO2、O3六个主要空气污染物作为指标，建立SOM网络模型，结果显示在空间分布上湖北省城市空气质量呈现从外围到中心降低的特征。采用数据挖掘中的关联规则，利用经典的Apriori算法挖掘主要空气污染物之间的关联关系，得到PM2.5、NO2、O3之间的强关联规则。

Abstract: With the development of economy and the increase of population, all kinds of environmental problems have taken place in our city. Firstly, taking PM2.5, PM10, SO2, CO, NO2, and O3 the six major air pollutants as indicators, self-organizing neural (SOM) network is used to study the urban air quality in Hubei province. The SOM neural network clustering model is established, and the results show that the air quality in Hubei province is reduced from the periphery to the center. Secondly, this paper uses the association rules in data mining and the classical Apriori algorithm to mine the correlation between PM2.5, NO2, and O3 the three major air pollutants, the strong association rules are found finally.