2019年基于深度信念网络的高压断路器故障识别算法.docxVIP

基于深度信念网络的高压断路器故障识别算法* 朱萌1，梅飞2，郑建勇1，沙浩源1，戴永正3，顾宇锋3 （1. 东南大学 电气工程学院，南京 210000；2. 河海大学 能源与电气学院，南京 211100；矚慫润厲钐瘗睞枥庑赖賃軔朧碍鳝绢懣硯涛镕頃赎巯驂雞虯从躜鞯烧。 3. 江苏南瑞泰事达电气有限公司，南京 210000） 摘要：针对高压断路器故障现有故障诊断算法中，特征提取不准确导致分类正确率较低的问题，提出了基于深度信念网络的高压断路器故障识别方法。深度信念网络（Deep Belief Network, DBN）是非监督的深度神经网络，由多个受限玻尔兹曼机（Restricted Boltzmann Machine, RBM）叠加起来组成。首先使用无标签的数据样本自下而上的对各RBM层逐层训练，得到各层最优参数；再以此为初始参数将DBN展开成反向传播的结构，使用带标签的数据样本进行全局的参数微调；最后得到DBN分类网络。这一过程中，有效避免了特征提取的人工操作，解决了网络训练的局部最优问题，使断路器故障诊断更加智能化。通过试验结果可知，该方法可准确、可靠地用于诊断断路器主要机械故障。聞創沟燴鐺險爱氇谴净祸測樅锯鳗鲮詣鋃陉蛮苎覺藍驳驂签拋敘睑绑。 关键词：断路器；故障诊断；深度信念网络；DBN；受限玻尔兹曼机 中图分类号：TM561 文献标识码：B 文章编号：1001-1390(2018)00-0000-00残骛楼諍锩瀨濟溆塹籟婭骒東戇鳖納们怿碩洒強缦骟飴顢歡窃緞駔蚂。 Fault recognition algorithm for high voltage circuit breakers based on deep belief network酽锕极額閉镇桧猪訣锥顧荭钯詢鳕驄粪讳鱸况閫硯浈颡閿审詔頃緯贾。 Zhu Meng1, Mei Fei2, Zheng Jianyong1, Sha Haoyuan1, Dai Yongzheng3, Gu Yufeng3彈贸摄尔霁毙攬砖卤庑诒尔肤亿鳔简闷鼋缔鋃耧泞蹤頓鍥義锥柽鳗铟。 (1. School of Electrical Engineering, Southeast University, Nanjing 210000, China. 2. College School of Energy and Electrical Engineering, Hohai University, Nanjing , 210000, China. 3. Jiangsu Nari Turbostar Electric Co.,Ltd., Nanjing, 210000, China)謀荞抟箧飆鐸怼类蒋薔點鉍杂篓鳐驱數硯侖葒屜懣勻雏鉚預齒贡缢颔。 Abstract: In order To to solve the problem of low classification accuracy due to incorrect feature extraction in the existing fault diagnosis algorithms for high voltage circuit breaker, the fault recognition algorithm based on deep belief network is proposed in this paper. Deep Belief belief Network network (DBN), as an unsupervised deep learning algorithm, is composed of multiple Restricted restricted Boltzmann Machines machines (RBM). Firstly, using the unlabeled data samples are used to train, each RBM layer is trained layer by layer from bottom to top, and the optimal parameters of each layer are obtained. Secondly, using the parameters obtained before are used as the initial parameters, the DBN is expanded into a structure that propagates backwards. Then, the labeled data samples are used for glo