Research On Fault Diagnosis Technology For Electric Power Communication Network

With the rapid development of the power communication network,its network topology becomes more and more complex,and it has more and more network devices,leading to the increase of fault types and the number of alarms in power communication network,and the correlation between alarm information is also more complicated.The current important research topics of fault diagnosis of power communication network is how to quickly and accurately find valuable information from a large number of alarm information to achieve the fault diagnosis.Based on the above,this paper mainly conducts the following research on the fault diagnosis technology of power communication network.1.The background and significance of the research are introduced,the research status of fault diagnosis technology in power communication network is analyzed,through the analysis,it can be seen that alarm information plays a decisive role in fault diagnosis.On this basis,the overall scheme design of fault diagnosis is accomplished.2.In order to achieve reliable and efficient diagnosis of power communication faults,it is necessary to preprocess massive fault alarm information to eliminate invalid,redundant and false alarm information.Therefore,this paper proposes a method of preprocessing alarm data based on clustering,sliding window and entropy method,by using density clustering to eliminate discrete alarms,using sliding windows to achieve alarm time synchronization and eliminate redundant alarm information,and finally the entropy method is used to weight the alarm information.The results of experiment shows that the method can effectively extract valuable alarm information and improve the efficiency of alarm information extraction.3.The traditional fault diagnosis technology of power communication network based on weighted FP-Growth(WFP)has the defects of low efficiency and low accuracy of fault diagnosis when processing a large amount of alarm data.Therefore,this paper proposes a weighted FP-Growth parallel power communication network fault mining method based on the Spark(WFPS).The simulation results show that the proposed method can complete the fault diagnosis of power communication network quickly and accurately.Finally,through the established simulation environment,by using the WFPS method and the WFP method to complete the fault diagnosis of the power communication network.The simulation results show that compared with WFP method,the WFPS method proposed in this paper has faster diagnosis response time,higher accuracy and lower fault missing rate in power communication network fault diagnosis.