微网故障检测关联规则挖掘算法研究  

Study on Association Rules Mining Algorithm for Micro-grid Fault Detection

作　　者：朱文灏[1,2] 郭其一[1] 

机构地区：[1]同济大学电子与信息工程学院,上海201804 [2]上海施耐德低压终端电器有限公司,上海201109

出　　处：《电气技术》2015年第8期7-10,共4页Electrical Engineering

摘　　要：针对于目前故障检测方法在微网应用中存在较大误差的问题,介绍了一种基于贝叶斯网络和关联规则数据挖掘的算法模型。首先,通过hash技术优化Apriori算法,对原数据挖掘,去除不期望的候选项集。然后,通过贝叶斯网络训练样本,减少检测误差,最终得到微网故障检测结果。仿真结果表明这种基于贝叶斯网络和关联规则挖掘算法的故障检测模型,比传统算法在配电网故障检测方面更有效率,并且检测误差大幅降低。In view of the problem that current fault detection methods exist large error in micro-grid fault detection, this paper presents a model based on Bayesian network and association rule mining. It firstly adopts Hash technology to optimize Apriori algorithm and remove the undesired candidate item set, conducts data mining of original data set, introduces Bayesian network for sample training to reduce detection error, and finally obtains power system detection result. Simulation results show that the proposed fault detection model based on Bayesian network and association rule mining is efficient in power system fault detection with detection error far less than that of traditional algorithm. Keywords.

关 键 词：关联规则挖掘 频繁项集优化 贝叶斯网络 微网故障检测 

分 类 号：TM769[电气工程—电力系统及自动化] TP311.13[自动化与计算机技术—计算机软件与理论]