Sensor signal analysis by neural networks for surveillance in nuclear reactors

The application of neural networks as a tool for reactor diagnosis is examined. Reactor pump signals utilized in a wear-out monitoring system developed for early detection of the degradation of a pump shaft are analyzed as a semi-benchmark test to study the feasibility of neural networks for monitoring and surveillance in nuclear reactors. The Adaptive Resonance Theory (ART 2 and ART 2A) paradigm of neural networks is used. The signals are collected signals as well as generated signals simulating the wear progress. The wear-out monitoring system applies noise analysis techniques and is capable of distinguishing these signals and providing a measure of the progress of the degradation. Results are presented of the analysis of these data, and the performances of ART 2-A and ART 2 for reactor signal analysis are evaluated