A New Hybrid Model CI-FPA-SVM For Prediction Of PM Daily Concentration In Kunming And Yuxi,China

With the rapid economic growth and fast expansion of urbanization,air pollution has already become extremely serious in China,especially for the particular matter.In order to solve the growing environment problems,Kunming and Yuxi,which are two important cities in Yunnan Province of China,are used as case studies to build model to predict daily pollution.In this paper,daily PM2.5 and PM10 concentration data form January 1,2015 to August 23,2016 are adopted to establish the novel hybrid model CI-FPA-SVM to predict the PM2.5 and PM10 concentration.The proposed model includes two parts: Firstly,due to its difficulty to access the possible correlation between different variables,the co-integration theory is introduced to get the input-output relationship.Then,the support vector machine(SVM),in which the parameters C and g are optimized by flower pollination algorithm(FPA),is used to obtain the nonlinear dynamical relationship between input and output.Four error indicators,including the root-mean-square error(RMSE),the mean absolute error(MAE),the mean bias error(MBE)and Pearson’s correlation coefficient(r)are adopted to evaluate the reliability of the new model CI-FPA-SVM.Compared with six benchmark models,FPA-SVM,CI-SVM,CI-GA-SVM,CI-PSO-SVM,CI-FPA-NN and multiple linear regression model,the new model CI-FPA-SVM is remarkably superior to all considered benchmark models for its high prediction accuracy.The empirical study results demonstrate that the application of the model for forecasting can give effective monitoring and management of air quality in the further.