采用自适应神经模糊推理系统的配电网故障分类方法  

Fault Classification Technique for Power Distribution Network Using Adaptive Network Based Fuzzy Inference System

作　　者：张钧[1] 李小鹏[1] 何正友[1] 

机构地区：[1]西南交通大学电气工程学院,四川省成都市610031

出　　处：《中国电机工程学报》2010年第25期87-93,共7页Proceedings of the CSEE

基　　金：国家自然科学基金项目(50877068);教育部优秀新世纪人才支持计划项目(NCET-06-0799)~~

摘　　要：配电网故障分类对配电网故障诊断、事故后分析具有重要作用。提出了一种基于自适应神经模糊推理系统(adaptive network-based fuzzy inference system,ANFIS)的中性点非有效接地系统的故障分类方法。利用小波变换提取故障特征频带内的暂态信号,基于统计量构造了用于故障分类的特征向量,研究了所构造的特征量在不同类型故障下的规律。利用自适应神经模糊推理系统,设计了一种用于小电流接地系统故障分类的方法。在PSCAD/EMTDC中建立了仿真模型,利用仿真样本对系统进行训练,测试样本的验证结果表明该方法具有较高的分类准确性。在中性点接地方式变化以及系统拓扑结构变化的情况下,研究了该方法的适应性,结果表明该方法的适应性良好。Fault classification is very important for fault diagnosis and post-fault analysis in power distribution systems. This paper proposes a fault classification technique for the neutral non-effectively grounded system using the adaptive network-based fuzzy inference system （ANFIS）. The transient signal in the frequency band with fault feature was extracted by using wavelet transform. Based on the statistic data, characteristic quantities for fault classification were calculated. Then the law of characteristic quantities was studied under different fault conditions. Based on ANFIS, a fault classification technique for the neutral non-effectively grounded system was proposed. A simulation model was established in PSCAD/EMTDC, and the training of classification system was conducted by utilizing simulating samples. Verification results of the test cases revealed that the proposed technique has a high veracity and a good adaptability under the condition of changes in the topology and the neutral grounding style .

关 键 词：配电网 小电流接地系统 故障分类 自适应神经 模糊推理系统 

分 类 号：TM76[电气工程—电力系统及自动化]