基于SVR的SDH光纤通信网络故障分类方法
Fault Classification Method for SDH Optical Fiber Communication Network Based on SVR
谭志强 1徐琳2
作者信息
- 1. 华东送变电工程有限公司,上海 200335
- 2. 国网上海青浦供电公司,上海 201799
- 折叠
摘要
同步数字体系(Synchronous Digital Hierarchy,SDH)光纤通信网络中的故障类型繁多,具体包括硬件故障、软件故障、配置错误以及光缆故障等.每种故障类型都有其独特的表现形式和原因,故障现象与原因之间的关联性难以确定,导致SDH光纤通信网络故障分类难度较大,为此提出基于支持向量回归(Support Vector Regression,SVR)的SDH光纤通信网络故障分类方法.基于泊松分布的故障数据传输稳态特征分布模型提取故障特征,结合故障特征向量与SVR分类SDH光纤通信网络故障.实验测试结果表明,该方法可以实现快速收敛,且分类准确率达到 99.0%以上,实际应用效果好.
Abstract
There are various types of faults in Synchronous Digital Hierarchy(SDH)fiber optic communication networks,including hardware faults,software faults,configuration errors,fiber optic cable faults,etc.Each type of fault has its unique manifestations and causes,and the correlation between fault phenomena and causes is difficult to determine,which increases the difficulty of fault classification in SDH optical fiber communication networks.Therefore,a fault classification method for SDH optical fiber communication networks based on Support Vector Regression(SVR)is proposed.The steady-state feature distribution model for fault data transmission based on Poisson distribution extracts fault features,and combines fault feature vectors with SVR to achieve fault classification in SDH optical fiber communication networks.The experimental test results show that this method can achieve fast convergence and a classification accuracy of over 99.0%,with good practical application effects.
关键词
支持向量回归(SVR)/同步数字体系(SDH)光纤通信网络/故障分类/统计方法/稳态特征分布模型/比特序列特征/特征向量Key words
Support Vector Regression(SVR)/Synchronous Digital Hierarchy(SDH)fiber optic communication network/fault classification/statistical method/steady-state characteristic distribution model/bit sequence features/eigenvector引用本文复制引用
出版年
2024
NETL
NSTL
万方数据