Abstract

. The intelligent approach to instrument surveillance and calibration verification described here was developed for use in nuclear power plants utilizing digitized data available in the Plant Computer. Data from Florida Power Corporations Crystal River #3 Nuclear Power Plant was recorded and applied to a five-layer autoassociative neural network (AANN) which is trained using a special robust training method. Plant variables that have some degree of coherence with each other constitute the inputs to the network. The relationships between the different variables are embedded in the neural network weights by the training process. As a result, the output can be a correct version of an input pattern that has been distorted by noise, missing data, intermittent faults, or non-linearities. Once the network has been trained with normal operational data using a robust training method, this system has been shown to successfully monitor the selected plant variables for sensor drift or failure by...

Citations