基于告警加权的智能光网络故障诊断算法

Automatic Switched Optical Network Fault Diagnosis Algorithm Based on Alarm Weighting

针对智能光网络故障诊断过程中多告警组合原因难以分析的问题,提出一种基于告警加权的故障诊断算法.通过保留频繁告警关联关系下所有告警间的置信度,采用基于大数定理的方法对各个告警事件进行加权处理,得到不同权重的告警事件,再在此基础上进行多告警组合下的原因分析,根据不同概率的原因组合进行故障诊断.实验证明,该算法显著提高了故障诊断有效性.

A fault diagnosis algorithm based on alarm weighting is proposed to solve the problem that the multi-alarm combination causes are difficult to analyze in the process of automatic switched optical network fault diagnosis. By retaining the frequent alarm correlation between all the alarm under the confidence level, the method uses the theorem of large number for weighting each alarm event, gets different weights of alarm events, and on the basis of many combinations of alarm analysis, fault diagnosis is done according to the reasons of different probability combination. Experimental results show that the algorithm improves the effectiveness of fault diagnosis.