| In recent years,the total carbon emissions of China presents a trend of continuously and rapidly rising.According to statistics,the total carbon emissions of China 5.513 billion in 2005,6.8 billion tons in 2008,7 billion tons in 2010,7.954 billion tons in 2012.It overall showed a rapid growth trend in carbon emissions,effective control of carbon emissions is become increasingly important.In the National Autonomous Contribution File,submitted in prior to the General Assembly in Paris,in 2005,China proposed that will make carbon dioxide emissions peaked in 2030 and strive to achieve it as soon as possible.Carbon dioxide emissions per unit of GDP in 2030 lower 60 to 65 percent than in 2005.In this context,by observing and analyzing carbon emissions factors in China,this article predict carbon emissions to gain carbon reduction strategies for China.In this paper,the prediction of carbon emissions and neural network technology will be closely linked.The carbon emissions forecasting model based on BP neural network,RBF neural network or Wavelet neural network are established.Focusing on the principle of neural network construction,training,predict and analysis of the results is aim to lay a solid theoretical foundation for the production of carbon emissions.Combined with collected historical data,BP neural network,RBF neural network and Wavelet Neural Network are respectively trained and emulated in MatlabR2014a.In order to achieve the desired fitting accuracy,iterative training approach is taken When samples being trained.By comparing all the effect of fitting,RBF neural network can be taken as the best prediction model of carbon emissions.According to the specific value of the carbon emission factors in the period of"the 13th Five-year".This paper take the model of carbon forecasting model based on RBF neural network to predict the carbon emissions from 2016 to 2020.And the carbon reduction strategies and suggestions will be raised in the end. |
