Predicting the Severity of Nuclear Power Plant Transients Using Nearest Neighbors Modeling Optimized by Genetic Algorithms on a Parallel Computer
- 1 July 1995
- journal article
- research article
- Published by Informa UK Limited in Nuclear Technology
- Vol. 111 (1), 46-62
- https://doi.org/10.13182/nt95-a35143
Abstract
The importance of automatic diagnostic systems for nuclear power plants (NPPs) has been discussed in numerous studies, and various such systems have been proposed. None of those systems were designed to predict the severity of the diagnosed scenario. A classification and severity prediction system for NPP transients is developed. The system is based on nearest neighbors modeling, which is optimized using genetic algorithms. The optimization process is used to determine the most important variables for each of the transient types analyzed. An enhanced version of the genetic algorithms is used in which a local downhill search is performed to further increase the accuracy achieved. The genetic algorithms search was implemented on a massively parallel supercomputer, the KSR1-64, to perform the analysis in a reasonable time. The data for this study were supplied by the highfidelity simulator of the San Onofre unit 1 pressurized water reactor.Keywords
This publication has 4 references indexed in Scilit:
- Application of Neural Networks to a Connectionist Expert System for Transient Identification in Nuclear Power PlantsNuclear Technology, 1993
- Abnormal Event Identification in Nuclear Power Plants Using a Neural Network and Knowledge ProcessingNuclear Technology, 1993
- Nuclear Power Plant Status Diagnostics Using an Artificial Neural NetworkNuclear Technology, 1992
- Nonlinear Identification of Process Dynamics Using Neural NetworksNuclear Technology, 1992