Yunnan Energy Demand Influence Factors Analysis And Prediction

Energy is an important material basis for economic development for one country.Since1978,the energy demand of the community kept growing along with the rapidgrowth of China’s GDP and in the late1990s, China became a net importer ofenergy,energy become the bottlenecks in economic development. Therefore,formulate scientific and rational energy development strategy can effectivelyguarantee that The stably and orderly development of the socio-economy. Energydevelopment strategies can not be separated from the energy impact factors analysisand forecasts of energy demand.Since2000, Yunnan provincial has made remarkable achievements on itssocio-economic development, while the energy consumption needs were increasing atan alarming rate. Even the energy production maintained a rapid growth, the supply ofthe energy products such as coal, electricity and oil is of a high degree of shortage.Yunnan Province has entered the medium-term of industrialization development,which also is a high energy-consuming stage, the demand for energy is grossing whilethe relative lack of resources in the province. Shortage of energy resources hasbecome a major bottleneck for the development of Yunnan Province. In thiscase, the effective energy demand forecasts is of great importance for energysecurity, energy investment, energy planning and the protection of economic andsocial development.There are lots of Energy demand forecasting methods. Some of them arefrequently used such as the ARMA model, trend extrapolation, the graysystem method, exponential smoothing, moving average method, the BPneural network prediction. Different prediction methods have their advantages anddisadvantages, they are not mutually exclusive, but are linked and mutuallyreinforcing. In view of this, Bates and granger proposed a combinationforecasting method. The Combination forecasting method absorbed the advantages of each individual prediction model, so it can effectively improve theprediction accuracy.This article used the regression model, the GM (1,1) model as well as threeexponential smoothing model for individual prediction. Then used the shapley valuein cooperative game to establish the weight of combination forecasting model forenergy demand forecast of Yunnan, and finely predicted the energy demand in thenext10years. The results of the combination forecasting model test showed that usingshapley value to determine the weight of factors in combination forecasting modelimproved the prediction accuracy, so, it’s an effective method to forecast energydemand.