Air Quality And Index Evaluation Analysis Of Handan City Based On Data Mining

The impacts of atmospheric environmental quality on human production and work are growing.With the rapid development of the economy,China's environmental problems are also facing tremendous pressure.In recent years,air pollution in some areas has been severe,especially small particles of high concentration air pollutants such as PM2.5 and PM10 and O3 near the ground layer frequently cause large-scale and long-lasting air pollution events.Therefore,research on air quality is very important and practical.Based on the air quality data of Handan from 2014 to 2017,this paper uses artificial intelligence algorithms to classify and predict its quality index and analysis air quality changes.Firstly,this paper conducts an annual time series analysis for the daily average air quality index AQI in Handan City.According to the statistical data and the trend chart,the air quality of each year is comprehensively compared.It is concluded that the air quality in the four years is low and flat in the middle,high and violent in the both heads.Second,in order to further analyze the air quality,this paper uses four clustering analysis methods to analyze whether the air quality changes have seasonal characteristics in Handan City.From the four clustering results,it is more appropriate to classify three categories,that is,winter is one type;autumn is one type;spring and summer are another type.Then based on the EMD empirical mode decomposition method,the air quality of Handan City is decomposed in time series,which verifies the seasonality and periodicity of air quality changes,and the air quality shows a trend of good development.Third,through the statistics of the first-ever air pollutants in the four-year period,PM2.5,PM10 and O3 ranked the top three.Finally,in the impact on air quality,we use the analytic hierarchy process to compute the weight of each pollutant.Among them,the specific gravity of sulfur dioxide is 0.0402,the specific gravity of carbon monoxide is 0.0314,the specific gravity of nitrogen dioxide is 0.0669,the specific gravity of PM10 is 0.2568,the specific gravity of ozone is 0.1489,and the specific gravity of PM2.5 is 0.4558.Fourthly,through the use of C5.0 decision tree,CART classification and regression tree,ANN artificial neural network and SPASS Molder based comprehensive model,the intelligent evaluation model of the city and air level is established according to air pollutants.According to the algorithm model,the air quality of Handan City is classified and predicted,and the prediction results are analyzed and summarized.It is concluded that the overall classification effect of C5.0 and comprehensive model is better.