摘要
离群数据挖掘是数据挖掘的一个重要内容,它为分析各种海量的、复杂的、含有噪声的数据提供了新的方法,但它在电力系统中还未得到广泛的应用。文中通过对现有的主要离群数据挖掘算法的简要对比说明,针对电力系统的基本特征提出应用信息熵原则的电力负荷离群数据挖掘改进算法,然后应用Kohonen网提取相关负荷的特征曲线,并将其用于不良数据的校正,通过对电力负荷的仿真分析表明了该算法的有效性。
Outlier data mining, as an important aspect of data mining, provides a new method for analyzing various quantitative, complex and noisy data, but is not widely used in power systems as yet. After a comparison of the existing outlier mining algorithms and based on the essential characteristics of electric loads, an improved outlier mining algorithm using entropy is proposed with the typical curve of correlating load extracted by Kohonen network to modify the bad data. Through the simulation of electric loads, the results show the effectiveness of the algorithm proposed.
出处
《电力系统自动化》
EI
CSCD
北大核心
2004年第11期41-44,86,共5页
Automation of Electric Power Systems
关键词
离群数据挖掘
负荷预测
聚类分析
信息熵
人工神经网络
outlier data mining
load forecasting
clustering analysis
entropy
artificial neural networks (ANN)