An Empirical Research Of Chinese No-production Energy Demand

It is well known that energy is the basic matter and dynamical resource for the development of economy and society. With China's rapid economic development in various trades and industries need more and more energy, at the same time, improving the level of national income, greater changes in consumption levels, which accompany the demand of No-production energy raise. At present, the strategy of sustainable development requires reasonably using energy, so energy-saving awareness of individual residents and formulating goods and building energy efficiency standards is significance for the country's developing policy "energy-saving emission reduction". Modeling and Anglicizing No-production energy demand is an important basis for learning the direction of Residents consumption in energy, formulating goods and building energy efficiency standards and planning transmission energy infrastructure construction layout.In this paper, firstly, analyzing the current situation and influencing factor on No-production energy demand for setting basement to model. Then taking the unstationary and non-linearity in the energy time series into consideration, the author models and analyzes China's No-production energy demand using methodology of cointegration, error correction and the neural network, putting forward two models of No-production energy demand that are reasonable, efficacious and can be used for prediction. Moreover, based on them, we make a demonstration study on the influential factors of No-production energy demand and then make a conclusion that there is a quantifiable relationship in the No-production energy demand and its influential factors. We can conclude that there exist a long-term equilibrium relationship between and its influencing factors by cointegration model. But by the neural network model, we can see that there is a non-linearity positive correlation relationship between No-production energy demand and consumption expenditure, urbanization rate. However, for the structure of energy consumption, there is a linearity positive correlation relationship. Secondly, we compare the two models (cointegration model and neural network one) and conclude that linear model (cointegration model) has vivid economic explaining power but lower forecasting precision while non-linear model(neural network model)has higher forecasting precision but unconspicuous economic explaining performance. Therefore, they can be complementary in application. Thirdly, we forecast the future No-production energy demand in China using the two models and then conclude that there is consistency in the forecasting result of two models. The results showed that the demand of Chinese No-production energy still have a high growth trend. Finally, we put forward some proposals about the development of No-production energy in China.