基于自忆性原理和EMD的中国清洁能源需求组合预测  

Combination Forecasting of Clean Energy Demand in China Based on Self-memorization Principle and Empirical Mode Decomposition

作　　者：汪哲荪[1] 袁潇晨[1] 魏一鸣[2] 金菊良[1] 焦建玲[3] 

机构地区：[1]合肥工业大学土木与水利工程学院,安徽合肥230009 [2]北京理工大学管理与经济学院,北京100081 [3]合肥工业大学管理学院,安徽合肥230009

出　　处：《水电能源科学》2010年第9期171-174,170,共5页Water Resources and Power

基　　金：国家自然科学基金资助项目(70941032,70733005)

摘　　要：以中国历年能源消费量为基础,分别建立灰色自忆性预测模型和数据机理自忆性预测模型;利用经验模态分解方法分析中国能源消费增长率在经济、人口、城市化下的变化情况,并建立基于经验模态分解的BP神经网络预测模型;通过遗传算法构建能源需求总量组合预测模型,求得中国清洁能源需求总量.研究结果表明:组合预测能够充分利用多个模型的丰富信息,提高预测的准确性;2020年中国清洁能源需求量将达到5.02×108～8.26×108 tce,应优先开发清洁能源.On the basis of energy consumption over the years in China,a gray self-memorization prediction model and a data based mechanism self-memorization prediction model have been separately developed.The variation of energy consumption growth rate in economy,population and urbanization has been studied by using empirical mode decomposition and a BP neural network prediction model based on empirical mode decomposition has been established.The total clear energy demand in China is obtained through the combination forecasting model based on genetic algorithm.The results show that the combination forecasting can take full advantage of information in each model and can improve the accuracy of prediction.Besides,clean energy should have the priority of developing because the demand of clean energy in China will reach 5.02×108~8.26×108 tce in the year of 2020.

关 键 词：自忆性 EMD 中国能源 清洁能源 能源需求量 组合预测模型 Empirical Mode Decomposition Based China Energy Demand CLEAN 神经网络预测模型 经验模态分解方法 需求总量 消费增长率 能源消费量 优先开发 遗传算法 变化情况 准确性 

分 类 号：TK79[动力工程及工程热物理—流体机械及工程] TV7[水利工程—水利水电工程]