A neural network based operation guidance system for procedure presentation and operation validation in nuclear power plants

Abstract

An operation guidance system (OGS) was developed to regulate and supervise operators' actions during abnormal environments in nuclear power plants (NPPs). The system integrated a primitive computerized procedures system (CPS) and ail operation validation system (OVS) imbedded in a virtual simulated operational environment. As the key component of the OGS, OVS provided two important functions for the operators: validated check of operations, and qualitative and quantitative effects analysis of operations. Each of operators' action was evaluated by the system and possible results were simulated by using artificial neural networks (ANN). Finally, corresponding suggestion or warning was provided to operators. This should reduce human errors during operation in emergency scenarios. (C) 2007 Elsevier Ltd. All rights reserved.