Identification of nuclear power plant transients with neural networks

Rapid identification of malfunctions is of premier importance for the safe operation of nuclear power plants. In order to provide sufficient lead time, malfunctions have to be identified within 60 seconds. A feedforward neural network trained with the backpropagation algorithm was developed to model simulated nuclear power plant malfunctions for a pressurized water reactor (PWR) and this model was then successfully applied to identify malfunctions of the Hungarian Paks nuclear power plant simulator.