Abstract
An operation support system for the experimental fast reactor "JOYO" was developed based on an intelligent alarm-handling. A specific feature of this system, called JOYCAT (JOYO Consulting and Analyzing Tool), is in its sequential processing structure that a uniform treatment by using design knowledge base is firstly applied for all activated alarms, and an exceptional treatment by using heuristic knowledge base is then applied only for the former results. This enables us to achieve real-time and flexible alarm-handling.The first alarm-handling determines the candidates of causal alarms, important alarms with which the operator should firstly cope, through identifying the cause-consequence relations among alarms based on the design knowledge base in which importance and activating conditions are described for each of 640 alarms in a frame format. The second alarm-handling makes the final judgement with the candidates by using the heuristic knowledge base described as production rules. Then, operation manuals concerning the most important alarms are displayed to operators.JOYCAT has been in commission since September of 1990, after a wide scope of validation tests by using an on-site full-scope training simulator.
Published Version (Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have