基于因散经验模式分解电力负荷混合预测方法.pdf

第32 卷 第8 期 电 网 技 术 Vol. 32 No. 8 2008 年4 月 Power System Technology Apr. 2008 文章编号：1000-3673 （2008 ）08-0058-05 中图分类号：TM715 文献标识码：A 学科代码：470·4054 基于因散经验模式分解的电力负荷混合预测方法 李媛媛，牛东晓，乞建勋，刘 达 （华北电力大学工商管理学院，北京市 昌平区 102206 ） A Novel Hybrid Power Load Forecasting Method Based on Ensemble Empirical Mode Decomposition LI Yuan-yuan ，NIU Dong-xiao ，QI Jian-xun，LIU Da （School of Business Administration，North China Electric Power University ，Changping District，Beijing 102206 ，China ） ABSTRACT: To solve the problem that at present common- 模式，因此能够克服依赖主观经验的缺点。然后，将这些内 used load analyzing methods mostly rely on subjective 在模式基于 fine-to-coarse 重构为高频、低频和趋势 3 个分 experiences and in classical empirical mode decomposition the 量。在对各分量特性进行分析的基础上，分别采用支持向量 mode mixing frequently appears, an ensemble empirical mode 机、自回归移动平均和线性回归模型对其进行预测。最后， decomposition (EEMD) based hybrid power load forecasting 将3 个分量的预测结果叠加作为最终的预测值。利用上述方 method is proposed. At first, by use of the improved algorithm 法对某电网进行24 点负荷预测，结果表明该方法可以有效 of empirical mode decomposition (EMD), i.e., the EEMD, the 地提高负荷预测精度。 power load series is decomposed, this way the objective series 关键词：因散经验模式分解；电力负荷；预测；内在模式； can be decomposed to several independent intrinsic modes 重构 adaptively, therefore the disadvantage of relying on subjective experiences can be overcome. Then, based on fine-to-coarse, 0 引言 these intrinsic modes are reconstructed as three components,