A new method for nuclear plant diagnostics using neural networks

摘要: This paper addresses nuclear power plant fault diagnosis using an artificial neural newtwork. In a previous work, single time snapshots of number variables were used for training the diagnostic network. this however, moving average variable is in attempt to reduce classification time. A range monitor (APRM) flux level build learning set on which backpropagation network trained. Preliminary efforts classify three transients by monitoring are presented.