| 摘要 |
<P>PROBLEM TO BE SOLVED: To predict an event such a device failure. <P>SOLUTION: The event prediction system includes an event occurrence model learning part 12, a status transition sequence calculation part 13, and a filtering module 14. In the event occurrence model learning part 12, based on an initial event occurrence model modeling a correlation between events which may occur in a monitoring target device and their event occurrence causes and transition between the event occurrence causes by using the finite automaton and a learning event sequence, a probability of occurrence of each event and a probability of transition between the event occurrence causes are calculated for each of the event occurrence causes, and an event occurrence model is generated. The status transition sequence calculation part 13 determines, based on the event occurrence model, a status transition sequence considered to have the highest probability of observing a failure cause detecting event sequence. The filtering module 14 predicts an event which may occur in the future in the monitoring target device, based on the event occurrence model and the status transition sequence determined by the status transition sequence calculation part 13. <P>COPYRIGHT: (C)2011,JPO&INPIT |
