基于粗糙集-贝叶斯方法的分布式电网故障诊断.pdfVIP

37 2 Vol.37, No.2 2010 3 　　　　　　　Journal of North China ElectricPower University 　　　　　　 Mar., 2010 - 1 1 2 　, , (1. , 071003; 2., 225009) :针对实际的电网故障诊断中常存在的冗余和不确定性信息这 一问题, 基于智 互补融合的思想, 将粗 糙集和贝叶斯有机结合在 一起, 并鉴于传统的人工智 在大电网故障诊断应用中存在的局限性而采用分布式 数据挖掘的思想, 提出了一种基于粗糙集-贝叶斯的分布式电网故障诊断方法。 该方法首先把大电网分割成 给定数目的局部电网, 然后利用粗糙集对各 个局部电网分别建立决策表并进行约简, 最后利用粗糙集-贝叶 斯方法进行局部电网内部及局部电网之间的联络线的规则提取形成全局规则库, 并根据实时报警信息进行规 则匹配得到诊断结果。 经电网故障诊断算例分析表明, 该算法正确、 有效, 速度快, 容错性好。 :大电网;故障诊断;粗糙集-贝叶斯;分布式数据挖掘 :TM711　　　:A　　　:1007-2691 (2010)02-0001-07 The distributed fault diagnosis of power networks based on Bayesian rough set ethod 1 1 2 LI Ran , GAO Cong-ying , ZHANG Lie-yong (1.School of Electrical and Electronic Engineering North China Electric Power University, Baoding 071003, China;2.Yangzhou Power Supply Co pany, Yangzhou 225009, China) Abstract:According to co ple entary strategy, a new power networks fault diagnosis ethod based on rough set and Bayes werepresented for fault diagnosis containing redundancy and uncertain infor ation.Further ore, consid- ering the li itations of traditionalartificial intelligence technology in thefault diagnosis of large scale power networks, distributed thinking was used to solve the proble .What atters ost is that thelarge scale power networks will be divided in a desired nu ber of sub-networks.What also counts is that decision table is established by using rough sets theory in these sub-networks.And additonal factor worth entioning is that the global rules base of sub-net- works and tielines is extracted by use of rough set and Bayes and the diagnosticrusults iso