| During the past two decades,some epidemiological studies have demonstrated that the major air pollutant impacts human health is particulate matter.PM2.5 is more likely to absorb hazardous and noxious substances.It is the carrier of all sorts of toxic substances in the air.However,because of the complexity of meteorological parameters and the difficult of quantitative,there exists a vast amount of uncertainty which causes PM2.5 concentration modelling predictions to differ from reality.In this paper,we are focusing attention on the forecasting of daily concentrations of PM2.5,which have been calculated in Lanzhou,China.According to the principle of “decomposition and ensemble”,we put forward a novel hybrid integrated model based on Ensemble Empirical Mode Decomposition(EEMD),artificial neutral networks(ANN)and Adaptive Particle Swarm Optimization(APSO)for PM2.5 concentration forecasting.The forecasting of daily PM2.5 concentrations is difficult because of the indeterminateness in describing meteorological problems.This hybrid integrated model is formulated exclusively to deal with difficulties in quantitating meteorological information,which has high volatility,irregularity and complicacy.We can decompose the hybrid integrated modal into three steps.Firstly,we utilize EEMD,a decomposition method,to decompose original data of PM2.5 concentration into a specific amount of independent intrinsic mode functions(IMFs).Afterwards,ANN with APSO which used to optimize the combination weights is applied to predict IMFs individually.Finally,another APSO-ANN is used to aggregate the predicted IMFs into a collection as the final forecasting results.The innovation of this novel hybrid integrated model is that it is effectively built on the principle of “decomposition and ensemble”,and the experimental results show that the forecasting results by the proposed hybrid integrated model are very accurate. |
