Research On Prediction Of Carbon Emissions In Hebei Province Based On Improved LSSVM Model

With the rapid development of economy and accelerating industrialization in China,energy consumption has increased substantially and the situation of carbon emission has been severe.It is an important measure to promote energy-saving and emission-reduction as well as develop low-carbon economy so that the sustainable development can be achieved.At present,the economic development in Hebei Province mainly depends on heavy industry,which leads to large energy consumption.As a typical province with high carbon emission,its haze and environmental pollution problems are highlighted,which pose a serious threat to the public’s health.Therefore,how to realize the coordinated development of economy,energy and environment has become a pressing problem in Hebei Province,and the trend of carbon emission in the future has turned into the focus of attention.Thus,Hebei Province is selected as the research object in this paper.The status and influential factors of carbon emissions are analyzed and the prediction is made for the next period,which has important practical significance for government to formulate related measures and promote carbon emission reduction in Hebei Province.In this paper,based on the data of energy consumption in Hebei Province from 1990 to 2015,the carbon emissions are calculated by the coefficient method and the overall level of carbon emissions is measured with three indicators,namely total carbon emissions,per capita carbon emissions and carbon emission intensity.Trough utilizing Tapio decoupling index,the relationship between economic growth and carbon emissions is explored and then the causal chain decomposition is carried out.The results show that there exists a weak decoupling in Hebei Province and the relationship is mainly determined by the elasticity of emission reduction.On this basis,from the structure of energy consumption structure,population,economy and technical level,the paper selects six influential factors including the proportion of coal consumption,population,urbanization rate,GDP,the proportion of the second industry and energy efficiency.The grey relational analysis(GRA)is employed to verify the correlation between carbon dioxide emissions and influential factors.The hybrid model GSC-LSSVM that incorporates cuckoo search algorithm based on Gauss disturbance(GCS)with least squares support vector machine(LSSVM)to study the relationship between carbon emissions and influential factors.Through taking the related data from 1990 to 2015 as an example,the good fitting precision of GS-LSSVM proves the feasibility and validity of this approach applied in carbon emission prediction.Considering the change of influential factors,this paper sets up three scenarios,namely natural scenario,low-carbon scenario and strong low-carbon scenario to forecast carbon emissions in Hebei Province from 2016 to 2025.According to the analysis of influential factors and prediction results of carbon emissions,the paper puts forward the relevant strategies of carbon emission reduction in Hebei Province,which provides theoretical reference for the formulation of policies on energy saving,emission reduction and low-carbon economy.