Study On Energy Demand Forecasting Model In Yunnan Province

Energy is an important resource which is essential to economic development, human survival and civilization. In future, China will complete the process of industrialization, urbanization and modernization. As a result, it will lead to a substantial increase in energy consumption. Therefore, predicting energy consumption and production scientifically. It has great significance for energy regulation and planning, promotion of Yunnan’s GRP and construction of a harmonious and resource-efficient society.Based on energy system theory and detailed analysis of energy consumption and production present situation in Yunnan, this paper predicts and analyzes energy consumption and production in Yunnan, finds out problems existing in Yunnan province energy system and makes some recommendations on development of Yunnan’s energy industry, comprehensive using multiple linear regression forecast model, GM (1,1) prediction model in grey theory, three exponential smooth method and vector autoregressive model.First, this paper discusses the relationship between energy problem and social coordinated development. Then this paper analyzes current situation of energy development and energy conservation pressure in Yunnan.Secondly, this paper introduces related concepts of energy system and content, classification, significance and major prediction methods of energy prediction.Thirdly, this paper comprehensively analyzes main factors affecting total energy consumption in Yunnan, establishes prediction models and inspects accuracy of prediction models.At last, this paper analyzes energy consumption characteristics and energy structure of Yunnan from aspects of energy production and energy consumption and finds out problems existing in Yunnan energy system. Based on construct energy prediction structure of Yunnan, analyze use of energy and formulate corresponding policies and regulations to provide institutional guarantee for development and effective utilization of energy industry.