Fractional Accumulation Grey Model And Its Application In Air Quality Prediction

With the rapid development of China's economy and the continuous acceleration of industrialization,the air pollution problem is increasingly prominent,especially the regional air pollution problem represented by the Beijing-Tianjin-Hebei region.Effective control of air pollution has become one of the urgent problems to be solved in the Beijing-Tianjin-Hebei region.Accurate prediction of the future air quality is an important prerequisite to effectively and quickly solve the air pollution problem,which is helpful to clarify the direction of air pollution control.In the same spatial dimension,the data of air pollutant concentration in different time dimensions vary greatly in their forms,indicating different air quality conditions.In this paper,four kinds of grey prediction models suitable for different data types are proposed,which are respectively used to predict the annual,quarterly,monthly and daily concentrations of air pollutants in the Beijing-Tianjin-Hebei region.The existing data on air pollutants concentration in the Beijing-Tianjin-Hebei region is few and the trend of data change is unstable.In view of this feature,the relevant literature on air quality prediction is sorted out,and it is found that the grey prediction model has a strong applicability and reliability for solving such problems.Based on the relevant theories of grey prediction model and the research status of air quality,fractional order accumulation operator is used to improve the grey prediction models that suitable for different data changes,and a series of fractional order grey prediction models are proposed,in which GM(1,1)model with fractional order accumulation(FGM(1,1))is used to predict the average annual concentration of each pollutant;Grey seasonal model with fractional order accumulation(FGSM(1,1))is used to predict the average quarterly concentrations of each pollutant;Seasonal GM(1,1)model with fractional order accumulation(FSGM(1,1))model is used to predict the average monthly concentration of each pollutant;The data grouping GM(1,1)model with fractional order accumulation(FDGGM(1,1))model is used to predict the daily concentration of each pollutant.In this paper,the modeling mechanism and steps of the above model are introduced in detail,and by comparing with other models,it is proved that the prediction ability of the proposed fractional-order model is superior.Based on the causes of air pollution and the current situation of air pollution control in theBeijing-Tianjin-Hebei region,combined with the prediction results of the model on the air pollutant concentration in each dimension,a series of suggestions are given for the long-term,short-term and coordinated control of air pollution in the Beijing-Tianjin-Hebei region.