Fault Diagnosis Orientedreseareh And Impiementa Tionof Paraileled Associate Rules

With the rapid development of information technology, devices becomeincreasingly integrated and complex. It has an important significance for detectingandpredicting failures promptly, ensuringthe efficiency and reliability of the equipmentwhen working, forming data with effective informationfrom the historical failuresprocessing. The traditionalfault diagnosis method has disadvantages such as thedifficultyof establishing diagnostic models, depending on the subjective experience,lackof accessrule, it is often powerless for the Diversity, Complexity, confidenceand theconnection between each other. Especiallyin the face oflarge datasets, it can notanalysisand processing efficiently. In this paper, deeply study of data mining association rules inparallel algorithms had been made against withthe traditional fault diagnosis of defects,and applied tofault diagnosis under way tocomplete the fault diagnosis of complexequipment with efficiency, preciseness and associated.Through the study of fault diagnosis system athome and abroad, and thediscussionon the application statusand development trends of fault diagnosis system, theinadequate of traditional fault diagnosis system in its ability of analysis and processinghad been discussed in this paper, thus put forward atechnology association rule of datamininginto thefault diagnosis system. Thebasic concepts andcore technologies ofcloudcomputing had been introduced, so did the keytechnologies such as HadoopDistributed File System Framework, MapReduce programming model, and so on. Thenwe introduced the concept, theobjects, the methodsand the applicationsin fault diagnosisof data mining. Then focuses on the theory and algorithm ofthe association ruleminingin data mining techniques, and propose an algorithm named Parallel Aprioriwhich based on MapReduce, which aimed at sparse characteristics of fault data,adopting the processing method of multi minimum support. Finally, based on thediscussionin the previous study, we integrate the parallel association rule algorithminfault diagnosis system, establish the correlation analysis for fault diagnosis subsystem,and describe thestructure and function ofthe subsystemin both aspects of design andimplementation. This paper has significance to some extend on using parallelassociation rule into fault diagnosis system.