Abstract
To prevent and mitigate an accident in nuclear power plant (NPP), it is important to identify the plant condition during the early stage of an accident. This research uses artificial intelligence techniques to check out the alarm status and to identify possible events. The Taiwan's Lungmen nuclear power station (LNPS), which is an advanced boiling water reactor (ABWR), is chosen as the target plant for the current study. A full scope engineering simulator is used to generate the testing data for method development. The following initiating events are considered in this study: small and large loss of coolant accidents (LOCAs), loss of feedwater events, closure of all main steam isolation valves, trip of all recirculation pumps, and loss of condenser vacuum. A time dependent look-up-table is established for each possible event. A pattern recognition algorithm is developed for fast tracking of the plant status. The developed algorithm is to serve as an early warning tool and to assist operator in cooperate with Containment Event Tree (CET) for mitigation of accidents.