小负荷短期预测中一种模糊聚类方法的探讨

作者：黄金榜1，吕 华2，吴国忠1 日期：2007-12-28/span> 浏览：3649 查看PDF文档

小负荷短期预测中一种模糊聚类方法的探讨
黄金榜1，吕 华2，吴国忠1
（1.浙江大学 电气工程学院,浙江 杭州 310027; 2.临安市供电局，浙江 杭州 311300)

摘 要：由于影响负荷的随机因素很多，小容量负荷受冲击负荷影响较大，传统模型中的相似日选择一般采取夹角余弦法，选取的精度有待提高。模糊聚类分析是模糊数学中一种有效的分析方法，应用模糊聚类分析中基于目标函数的方法可对负荷的历史样本进行高精度的分类，所采用的HCM算法有效地提高了分类精度；在此基础上运用AGO回归算法进行合理的数据拟合后得到预测结果。算例证明，此方法是十分有效的，提高了小负荷短期负荷预测的精度。
关键词：模糊聚类;硬C划分;累加回归;小容量负荷

An fuzzy clustering method for shortterm small load forecasting
HUANG Jinbang1, LV Hua2, WU Guozhong1
(1.College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China;
2.Lin’an Municipal Power Supply Bureau, Hangzhou 311300, China)

Abstract: As a result that the load was influenced by many random factors and the small load was greatly influenced by load pulses. And the precision of cosineangle method for selecting similar day was not very well. The fuzzy clustering classifying method was applied and the precision of classification was greatly improved. Based on the above, the AGO regressive model was used properly and sufficiency. The example shows that the method is efficiency, and it improves the precision of small load forecasting greatly.
Key words: fuzzy clustering; hard cmeans (HCM) clustering; AGO regressive; small load

参考文献(Reference)：
［1］RO D. Pattern Classification and Scene Analysis［M］. NewYork: John Wiley & Sons Inc.,1973.
［2］GERMOND A. Heterogeneous artificial neural network for short term electrical load forecasting［J］. IEEE Transactions on Power Systems,1996,11(1):397-402.
［3］SLOBODAN R. Weather sensitive method for short term load forecasting in electric power utility of serbia ［J］. IEEE Transactions on Power Systems,2003,18(4):1581-1586.
［4］高新波.模糊聚类分析及其应用［M］.西安：西安电子科技大学出版社，2004.
［5］张贤达.现代信号处理：第2版［M］.北京:清华大学出版社,1996.
［6］飞思科技产品研发中心.神经网络理论与Matlab 7实现［M］.北京：电子工业出版社，2005.
［7］冯 定.神经网络专家系统［M］.北京：科学出版社，2006.
［8］杨伦标.模糊数学原理及应用［M］.广州：华南理工大学出版社，1992.