Abstract: Generally, the monitoring system for power generation equipments adopts absolute value to issue alarms when the operating parameters exceed the set values. This single monitoring method makes it difficult to notice the early signs of the equipment abnormalities and then keep the track of their further development, which may eventually cause the forced outage of equipments. In this paper, a modeling technique based on similarity mechanism(SBM) is proposed in which a model matrix is established to simulate the real equipments by analyzing the actual operation data with mathematical method. Then, the deviation is obtained by comparison between the actual operation value and the calculated value on the model. If it is more than the preset deviation, the alarm will be triggered. This system can display the online dynamic process of the system parameter changes, reflect how well the system is operated, and timely remind the maintenance personnel of equipment maintenance. At last, the practical applications of the system are demonstrated with case studies.

Key words: power generating equipment, state mornitoring, similarity-based modeling (SBM), data mining, failure prognostic