中长期电力负荷预测模型筛选与组合方法  

Sifting and Combination Method of Medium-and Long-Term Load Forecasting Model

作　　者：金鑫[1] 罗滇生[1] 孙广强 张红旭 郑盾[3] 武斌[3] 李朝阳[3] 

机构地区：[1]湖南大学电气与信息工程学院,长沙410082 [2]河南省电力勘测设计院,郑州450007 [3]郑州供电公司,郑州450006

出　　处：《电力系统及其自动化学报》2012年第4期150-156,共7页Proceedings of the CSU-EPSA

摘　　要：将预测决策思想引入到中长期电力负荷组合预测中,并深入分析了现有预测模型筛选方法的优缺点,在此基础上,完善了以灰色关联度和预测有效度两类指标为基础的预测模型评价指标体系。依据提出的协调因子、综合有效性指标等概念,构造了基于综合有效性指标和模型冗余校验的预测模型筛选方法和基于综合有效性指标体系的中长期电力负荷变权组合预测模型。算例表明,文中提出的组合预测模型预测精度较高,实用性强。The idea of forecasting and decision--making is introduced for multi-model combination in this paper. Based on the merit and demerit of the current forecasting model screening methods the evaluation system of indicator for forecasting model on the basis of grey relational grade and forecasting validity is improved..Ac- cording to the concepts proposed previously, such as the coordination factor and comprehensive validity index, the forecasting model screening method based on the comprehensive validity index and model redundancy verifi- cation is established. The medium-and long-term load variable weighting combination forecasting model based on the comprehensive validity index system is also presented. The proposed model is proved to be more effec- tive and accurate from the analysis of the instances.

关 键 词：组合预测 中长期负荷预测 综合有效性指标 协调因子 

分 类 号：TM715[电气工程—电力系统及自动化]