基于数据挖掘多层次细节分解的负荷序列聚类分析.pdfVIP

30 卷 2 期 电 网 技 术 Vol. 30 No. 2 2006 年1 月 Power System Technology Jan. 2006 文章编号：1000-3673 （2006 ）02-0051-06 中图分类号：TM715 文献标识码：A 学科代码：4704051 基于数据挖掘多层次细节分解的负荷序列聚类分析 张智晟，孙雅明，张世英，赵 艳 （天津大学，天津市 南开区 300072 ） Clustering Analysis of Electric Load Series Using Clustering Algorithm of Multi-Hierarchy and Detailed Decomposition Based on Data Mining ZHANG Zhi-sheng，SUN Ya-ming，ZHANG Shi-ying，ZHAO Yan （Tianjin University，Nankai District，Tianjin 300072 ，China ） ABSTRACT: A load clustering algorithm (CA) using 差分序列均方差；Kohonen神经网络；短期负荷预测 multi-hierarchy and detailed decomposition based on data 1 引言 mining (DM) and its performance evaluation index are proposed, in which the Euclidean distances and mean square [1-7] 电力系统短期负荷预测 （Short-Term Load deviation of difference series of load series are used to Forecasting，STLF ）是电力系统正常运行、安全调 construct optimal criterion of intersection, at the same time the multi-hierarchy and detailed decomposition clustering is 度操作、电力市场运营的重要依据，准确的负荷预 controlled by adding requirements of corresponding parameters 测有助于降低实时电价。 according to the sensitivity of random factors to loads, and it is 由STLF的研究可知，负荷内部的复杂非线性使 an important base for improving load forecasting precision to 负荷序列具有混沌特性，必须依据混沌理论及动力 cluster the detailed extent of load curve s contour similarity.