Research On Fault Diagnosis Of Transmission Equipment In Communication Network Management System

Communication network as an important part of the power system,effectively protect the power production,dispatch,relay protection and other communications business needs.In order to ensure the normal operation of the power system,faults must be diagnosed in the shortest possible time in the event of a fault,so as to avoid incalculable losses.However,the relevance of communication networks is rather complicated.A single failure often triggers a series of alarms,and not all alarms clearly indicate the root cause of the alarm.Because these alarms have a certain relevance,we can conduct correlation analysis and dig out the potential association rules to provide a strong basis for fault diagnosis.This paper first introduces the research background and significance of the research,the fault diagnosis of communication network management system and the research status of alarm correlation analysis,and then elaborates the definition and classification of communication network and the transmission equipment SDH in communication network management system.In addition,the concept of alarm correlation analysis and data mining is introduced.At last,aiming at the shortcomings of Apriori algorithm,this paper points out the algorithm of association rules mining based on multi-minimum support-MS-Apriori,and optimizes the statistics of candidate itemset support number and generating association rules,avoid repeated scanning of the database to improve the efficiency of the algorithm.In addition,an algorithm for post-processing rules to reduce redundancy is given.In this paper,the Apriori algorithm and the optimized MS-Apriori algorithm are simulated.Firstly,the alarm data is preprocessed and the transaction packet is completed by WINEPI algorithm.Then,the Apriori algorithm is implemented by mining structure,and different properties of alarm are mined,found that the alarm named as the keyword mining results more conducive to analysis of root alerts.Finally,Apriori and optimized MS-Apriori are implemented respectively.The mining results and execution efficiency of the two algorithms are compared,and the redundant algorithm is used to eliminate the redundancy.Finally,the alarm correlation analysis of SDH in a company’s power communication network management system is completed,the root alarm is diagnosed and the location is realized.