基于组合预测模型的云南省能源消费预测研究

Study on Energy Consumption Prediction in Yunnan Province Based on Combined Prediction Model

摘要: 能源问题关系着我国经济的发展、环境的保护以及人类的健康，准确预测能源消费总量至关重要，可以给能源工作决策提供科学依据。本文结合云南省能源消费总量的历史数据，在三个单一预测模型的基础上，基于均方误差的非最优权重线性组合法和最小预测误差平方和为目标函数的最优权重线性组合法构建云南省能源消费总量的组合预测模型。通过比较上述预测结果，发现组合预测模型具有更好的预测效果，研究表明该模型对云南省能源消费预测有重要的理论与现实意义。

Abstract: There is a strong relationship between energy and economic development of China, the protection of human health and the environment. It is essential to accurately predict energy consumption. This can provide a scientific basis for decision-making about energy work. Based on the historical data of energy consumption in Yunnan Province, on the basis of the three single prediction model, we establish combined prediction model to make energy predictions in Yunnan Province by non- optimal weighted linear combination model and minimum sum of square error of objective func- tion for optimal weight linear combination model. The prediction results of these models are ana- lyzed and compared. The results show that combined prediction model is performed better. This indicates that the combined prediction model is an useful theoretical tool for energy prediction in Yunnan Province.