Neural networks for fault diagnosis of a nuclear fuel processing plant at different operating points

摘要: Abstract Industrial processes often produce at various operating points; however, demonstrated applications of neural networks for fault diagnosis usually consider only a single (primary) point. Developing standard neural-network scheme all points may be impractical due to the unavailability suitable training data less frequently used (secondary) points. This paper investigates application non-catastrophic faults in an industrial nuclear processing plant different Data-conditioning methods are investigated facilitate classification, and reduce complexity networks. Results illustrate performance trained classifying process using simulated real data.