Research On The Method Of Fault Predicting Of Steam Generator

The Steam Generator is one of the most important components of Nuclear Power Plants. Therefore it is very significant that preparing the fault prediction systems to the Steam Generator, which can ensure the Steam Generator to operate safely and reduce the rate of accidents.Fault monitoring and Fault diagnosing are new scientific researching fields developed in these years. People want to get the information of the fault before it makes the harm serious and obvious. So we put forward fault prediction which is a new science research field.A lot of national and international fault prediction and fault diagnosis systems for the nuclear power plant are studied in the paper and set as basis of the research. In the paper, a fault predicting method for the Steam Generator, which is based on the predictive BP neural network trained using time series data. By using the error researching between practical data and predictive data the fault prediction is realized. At last, we prove the method can be used for the fault of Steam Generator.The thesis finally utilizes Visual C++ platform to set up a Fault Prediction System of Steam Genetic based on the predictive BP neural network used time series dates. It adopt human-computer interaction technologies interface and has the characteristic of friendship interface and simple operation.