Research On Fault Diagnosis Model For The Vacuum Of Condenser

The quality for the operating state of condenser has a great influence on thesafety and economy of the plant. In the operating process of unit, the internalcondenser will continue to leak into the air, it will cause the vacuum degree ofcondenser to drop and produce various faults in the system. Therefore the staff hopeto get timely treatment failure in the actual power plant, so as to ensure the unit towork safely and reliably. Therefore, the establishment of model for monitoring andfault diagnosis for condenser system to improve the unit economic operation level hasimportant theory meaning and practical value.This paper has a deep analysis of condenser running characteristics, analysis ofthe impact of the main factors from the theoretical and practical in vacuum, usingprincipal component analysis and particle swarm optimization BP neural networkcombination method, established forecast a model of the condenser vacuum system topredict the vacuum value, it can monitor the units run. And by calculating thecondition of vacuum in condenser is the optimal value, operator will adjust operatingparameters, which can ensure the crew in the best working condition.Because condenser’s faults have many reasons, this article through the massiveliterature summary and the power plant real data establishes a condenser faultdiagnosis knowledge base, combination of the neural network with expert system tobe realized the fault diagnosis of condenser. So this paper paste development themonitoring vacuum of the condenser and fault diagnosis system software, it canachieve on-line monitoring vacuum state and fault diagnosis. This system not onlyhas the advantages of simple operation, but also can directly get the diagnosis resultsand has a guiding role for unit operation personnel.