摘要
该文提出了基于面向元件神经网络与模糊积分信息融合技术的电网故障诊断方法,首先针对单个线路、母线、变压器设备分别建立面向元件的神经网络模型,以面向元件神经网络作初步诊断,在初级诊断的基础上,结合电网拓扑关系,应用模糊积分信息融合技术进行综合诊断。最后通过算例测试验证了其有效性。该方法克服了获取训练样本和适应拓扑变化的问题,提高了诊断准确率,对电网复杂故障有较好的诊断能力。
This paper presents a novel diagnosis method combining element-oriented artificial neural networks and fuzzy integral fusion. The proposed method models the transmission line,bus and transformer using element-oriented ANNs. When a fault occurs,a primary diagnosis is made by element-oriented ANNs,and then the synthetic diagnosis fuses the primary diagnosis results employing fuzzy integral. The proposed method overcomes the Achilles heel of ANNs at getting training patterns and handling topology changes. And the simulation shows that by the use of synthetic diagnosis,the accuracy of diagnosis system is effectively improved. This method is promising for application in large scale real-time fault diagnosis.
出处
《电工技术学报》
EI
CSCD
北大核心
2010年第9期183-190,共8页
Transactions of China Electrotechnical Society
基金
国家自然科学基金(50677062)
新世纪优秀人才支持计划(NCET-07-0745)
浙江省自然科学基金(R107062)
国家863计划(2008AA05Z210)资助项目
关键词
电力系统
故障诊断
面向元件神经网络
综合诊断
模糊积分
信息融合
Power systems
fault diagnosis
element-oriented artificial neural networks
synthetic diagnosis
fuzzy integral
information fusion
