期刊
PROGRESS IN ELECTROMAGNETICS RESEARCH-PIER
卷 136, 期 -, 页码 457-478出版社
EMW PUBLISHING
DOI: 10.2528/PIER12121402
关键词
-
资金
- National Natural Science Foundation of China [61107051]
- National High Technology Research and Development Program of China [SS2012AA010407]
- Major Program of Natural Science Foundation of Hubei Province [2011CDA114]
Based on Time-Domain Reflectometry (TDR) technique, a novel method which could locate faults on the coaxial cable distribution network by using Support Vector Machine (SVM) is proposed in this paper. This approach allows the faulty network to be reconstructed by estimating the lengths of branches. A State-transition Matrix model is employed to simulate the TDR response at any port and evaluate the transfer function between two points. SVM is used to solve the inversion problem through training datasets created by the State-transition matrix model. Compared to the existing reflectometry methods, our proposed method can tackle multiple faults in the complex cable networks. Numerical and experimental results pointing out the performance of the SVM model in locating faults are reported.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据