正在加载图片...
第9卷第2期 智能系统学报 Vol.9 No.2 2014年4月 CAAI Transactions on Intelligent Systems Apr.2014 D0I:10.3969/j.issn.1673-4785.201310026 网络出版地址:http://www.cnki.net/kcms/doi/10.3969/j.issn.1673-4785.201310026.html 压缩感知理论中的建筑电气系统故障诊断 张龙,陈宸2,韩宁1,王亚慧3 (1.北京林业大学工学院,北京100083;2.美国德州大学Dallas分校电子工程系,Richardson75080:3.北京建筑大 学电气与信息工程学院,北京100044) 摘要:针对目前建筑电气系统缺少有效诊断故障方法的问题,提出一种基于压缩感知理论的建筑电气系统故障分析 诊断方法,其中的关键是将故障的分类归结为一个求解待测样本对于整体训练样本的稀疏表示问题。使用建筑电气故 障模拟平台数据建立了其故障诊断模型,分别采用支持向量机、1,分类器和L,分类器对系统的5种故障状态进行了诊 断对比,结果表明,利用稀疏表示算法可以达到很好的诊断效果,分类准确率为96.4%,诊断运行时间0.2601s,可以满 足小样本数据的建筑电气故障诊断工程应用的需求。 关键词:电气故障:故障诊断:支持向量机:压缩感知:分类:特征提取:信号重构:最优化 中图分类号:TP18;TM743文献标志码:A文章编号:1673-4785(2014)02-204-06 中文引用格式:张龙,陈宸,韩宁,等.压缩感知理论中的建筑电气系统故障诊断[J].智能系统学报,2014,9(2):204-209。 英文引用格式:ZHANG Long,CHEN Chen,HAN Ning,etal.Fault diagnosis of electrical systems in buildings based on com- pressed sensing[J].CAAI Transactions on Intelligent Systems,2014,9(2):204-209. Fault diagnosis of electrical systems in buildings based on compressed sensing ZHANG Long',CHEN Chen2,HAN Ning',WANG Yahui' (1.School of Technology,Beijing Forestry University,Beijing 100083,China;2.Department of Electrical Engineering,University of Texas at Dallas,Richardson 75080,China;3.Department of Electrical and Information Engineering,Beijing University of Civil Engi- neering and Architecture,Beijing 100044,China) Abstract:In order to diagnose a fault in building electrical systems effectively,this paper presents a new fault a- nalysis and diagnosis method based on the compressed sensing theory.The key is to boil down the fault classification into a problem of representing a testing sample as a sparse linear combination of the training samples.A fault diag- nosis model was established by using the building electrical fault data simulated from the hardware experimental platform.Five different fault statuses of the system were diagnosed by using the support vector machine,classifierl and classifierl,respectively.The experimental results showed that our method of using sparse representation can a- chieve good diagnostic results.The accuracy rate of the classification was 96.4%,the operation time of the diagnosis was 0.260 1 s,and therefore,the method meets the application demands for the diagnosis of a building electrical fault with small specimen data. Keywords:electric breakdown;fault diagnosis;support vector machines;compressed sensing;classification;fea- ture extraction;signal reconstruction;optimization 随着城市化进程的加速,高层和超高层建筑日益 增加,人们对于建筑物安全和舒适度的要求也越来越 高。在整个建筑物中,建筑电气是关键技术之一,它 收稿日期:2013-10-15.网络出版日期:2014-03-31. 包括了照明系统、供配电系统、动力设备系统、办公及 基金项目:北京市自然科学基金资助项目(8111002) 通信作者:张龙.E-mail:long1988iacf@163.com. 管理自动化等主要内容。不同子系统间的相互关联第 9 卷第 2 期 智 能 系 统 学 报 Vol.9 №.2 2014 年 4 月 CAAI Transactions on Intelligent Systems Apr. 2014 DOI:10.3969 / j.issn.1673⁃4785.201310026 网络出版地址:http: / / www.cnki.net / kcms/ doi / 10.3969 / j.issn.1673⁃4785.201310026.html 压缩感知理论中的建筑电气系统故障诊断 张龙1 ,陈宸2 ,韩宁1 ,王亚慧3 (1. 北京林业大学 工学院,北京 100083; 2. 美国德州大学 Dallas 分校 电子工程系,Richardson 75080; 3. 北京建筑大 学 电气与信息工程学院,北京 100044) 摘 要:针对目前建筑电气系统缺少有效诊断故障方法的问题,提出一种基于压缩感知理论的建筑电气系统故障分析 诊断方法,其中的关键是将故障的分类归结为一个求解待测样本对于整体训练样本的稀疏表示问题。 使用建筑电气故 障模拟平台数据建立了其故障诊断模型,分别采用支持向量机、 l 1 分类器和 l 2 分类器对系统的 5 种故障状态进行了诊 断对比,结果表明,利用稀疏表示算法可以达到很好的诊断效果,分类准确率为 96.4%,诊断运行时间 0.260 1 s,可以满 足小样本数据的建筑电气故障诊断工程应用的需求。 关键词:电气故障;故障诊断;支持向量机;压缩感知;分类;特征提取;信号重构;最优化 中图分类号: TP18; TM743 文献标志码:A 文章编号:1673⁃4785(2014)02⁃204⁃06 中文引用格式:张龙,陈宸,韩宁,等. 压缩感知理论中的建筑电气系统故障诊断[J]. 智能系统学报, 2014, 9(2): 204⁃209. 英文引用格式:ZHANG Long, CHEN Chen, HAN Ning, et al. Fault diagnosis of electrical systems in buildings based on com⁃ pressed sensing[J]. CAAI Transactions on Intelligent Systems, 2014, 9(2): 204⁃209. Fault diagnosis of electrical systems in buildings based on compressed sensing ZHANG Long 1 , CHEN Chen 2 , HAN Ning 1 , WANG Yahui 3 (1. School of Technology, Beijing Forestry University, Beijing 100083, China; 2. Department of Electrical Engineering, University of Texas at Dallas, Richardson 75080, China; 3. Department of Electrical and Information Engineering, Beijing University of Civil Engi⁃ neering and Architecture, Beijing 100044, China) Abstract: In order to diagnose a fault in building electrical systems effectively, this paper presents a new fault a⁃ nalysis and diagnosis method based on the compressed sensing theory. The key is to boil down the fault classification into a problem of representing a testing sample as a sparse linear combination of the training samples. A fault diag⁃ nosis model was established by using the building electrical fault data simulated from the hardware experimental platform. Five different fault statuses of the system were diagnosed by using the support vector machine, classifier l 1 and classifier l 2 , respectively. The experimental results showed that our method of using sparse representation can a⁃ chieve good diagnostic results. The accuracy rate of the classification was 96.4%, the operation time of the diagnosis was 0.260 1 s, and therefore, the method meets the application demands for the diagnosis of a building electrical fault with small specimen data. Keywords: electric breakdown; fault diagnosis; support vector machines; compressed sensing; classification; fea⁃ ture extraction; signal reconstruction; optimization 收稿日期:2013⁃10⁃15. 网络出版日期:2014⁃03⁃31. 基金项目:北京市自然科学基金资助项目(8111002). 通信作者:张龙. E⁃mail:long1988iacf@ 163.com. 随着城市化进程的加速,高层和超高层建筑日益 增加,人们对于建筑物安全和舒适度的要求也越来越 高。 在整个建筑物中,建筑电气是关键技术之一,它 包括了照明系统、供配电系统、动力设备系统、办公及 管理自动化等主要内容。 不同子系统间的相互关联
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有