An Intelligent Monitoring System for Electric Power Variation in a Nuclear Power Plant

This paper presents an electric power monoring based on Artificial Neural Network (ANN) for the nuclear power plants. The Recurrent Neural Networks (RNN) and the feed-forward neural network are selected for the plant modeling and anomaly detection because of the high capability of modeling for dynamic behaviors. Two types of Recurrent Neural Networks (RNN) are used. The first one Elman type of RNN which has a feed-back from hidden layer to the input layer neurons while in the Jordan type, from the outputs of the neural net to the inputs of the neural net. Although this approach enables to realize the whole system condition monitoring in operating nuclear power plant (NPP), we are especially focused on active power and reactive power monitoring as well as power factor monitoring. Today, competition in electric power supply industry needs to properly evaluate plant capabilities and produced power quality, so the monitoring of active and reactive power becomes an important issue. Therefore active and reactive power and their variations are monitored taking their signals using the assigned channels and the electric power coefficient is simultaneously monitored from these measured reactive and active electric power signals.