Research On The Direct Carbon Emission Forecasting System Of China's Provincial Residents Based On Neural Network

Global climate change is closely related to human carbon emissions and will affect the regional economic and social development,natural ecological environment,food security and water supply and so on.Global climate system is extremely complex,the factors which leading to global climate change are complex and diverse,it has become one of the important issues of common concern in today's society.The fourth assessment report issued by the Intergovernmental Panel on Climate Change(IPCC)attributes the causes of global warming to Greenhouse Gas emissions from human activities,and further confirms that the conclusion is scientific in the fifth assessment report.As a consensus among governments on climate change,this conclusion provides an important theoretical basis for countries to address global change.In recent years,it has a fierce game round the carbon emissions obligations among developed countries and developing countries and underdeveloped countries.Today,China is one of the largest developing countries in the world,and also is one of the major energy consuming countries.China should formulate scientific and fair carbon emission plans considering its historical process and realistic development,as well as fulfilling the obligations of responsible powers.On November 30,2015,the Chinese President Xi Jinping delivered a speech and promised that China's carbon emissions will reach its peak around 2030 at the opening ceremony of the Paris Conference on Climate Change in Paris,France.Therefore,it will have important policy theory and guiding significance to forecast the direct carbon emissions of China's provincial residents,which according to the current situation of China's energy consumption and the direct level of carbon emissions about Residents of provinces and cities.Based on the above background,this paper collected 19 kinds of energy consumption situationsfrom 1995 to 2014 in China's 30 provinces and autonomous regions(the data of Tibet,Hong Kong,Macao and Taiwan is missing),and accounted out the direct consumption of carbon emissions data of the million people per capita about 30 provinces and autonomous regions,which is according to the total population of various provinces.On the basis of summing up previous theories and research progress,the author developed to achieve the direct carbon emission forecasting system of Chinese province residents based on the method of GIS and neural network model.The main research contents and innovations are as follows.(1)It mainly expounded the current situation of China's carbon emission and the facing pressure of emission reduction,reviewed and summarized the research progress of the domestic and foreign scholars on carbon emissions accounting and forecast,as well as neural network model application.Then,the feasibility prediction of the neural network model using to forecast the direct carbon emissions by provincial residents based on Matlab.(2)It organized and analysed the energy statistics yearbook and the demographic data of every year and the provincial and municipal,and obtained China's provinces direct carbon emissions data and per capita carbon emissions data from 1995 to 2014 by using the carbon emission coefficient method and mathematical statistics and other accounting methods,which provides a data base for the prediction of direct carbon emissions data for residents.(3)according to the characteristics that neural network model is suitable for dealing with the strong non-linearity of the data,three kinds of intelligent prediction models are chose,such as BP neural network,RBF neural network and Elman neural network.It mainly analysis the structural principle,learning algorithm and design flow of the models and predicts the direct carbon emissions of residents in Beijing.And the optimal model is selected by the comparision the prediction accuracy of the three methods.The result shows that the accuracy of the prediction results by Elman neural network model is higher thanothers and more suitable for carbon emission data projections.(4)It is main to describe the system requirements analysis and its design principles,and build the system function modules using VS2012,ArcGIS Engine10.2 and Matlab R2014 a and other software platforms.The Elman neural network model in Matlab is embedded into the system,and the C# language design is used to realize the direct carbon emission forecasting system of Chinese province residents.The system mainly includes the basic map operation,data management,map layout,carbon emission thematic map production and carbon emission prediction module based on Elman neural network,and the test runs stable.(5)the development of intelligent forecasting system is used to realize the direct carbon emission forecast for Chinese provinces in 2015-2035.The result shows that the growth of direct carbon emissions of the million people per capita in China after 2020 is relatively slow and will reach a relatively stable peak in 2032 with the situation of absence external forces and residents of direct carbon emissions free development.This is also confirmed that it is achievable that China's commitment to the carbon emissions will reach the peak before 2030.In a word,this paper analyzes the feasibility of BP,RBF and Elman neural network model for predicting the carbon emission of residents,and it integrated visualization intelligent platform that easy to operate,stable and scalable,which based on the theory of carbon emission and the theory of neural network model and with the support of GIS component technology and intelligent method.Then the trends and characteristics of 2015-2035 were analyzed and analyzed to provide scientific basis for the control and planning of carbon emissions based on the direct carbon data of Chinese provinces from 1995 to 2014.