Predicting the Severity of Nuclear Power Plant Transients Using Nearest Neighbors Modeling Optimized by Genetic Algorithms on a Parallel Computer

摘要: The importance of automatic diagnostic systems for nuclear power plants (NPPs) has been discussed in numerous studies, and various such have proposed. None those were designed to predict the severity diagnosed scenario. A classification prediction system NPP transients is developed. based on nearest neighbors modeling, which optimized using genetic algorithms. optimization process used determine most important variables each transient types analyzed. An enhanced version algorithms a local downhill search performed further increase accuracy achieved. was implemented massively parallel supercomputer, KSR1-64, perform analysis reasonable time. data this study supplied by high-fidelity simulator San Onofre unit 1 pressurized water reactor.