A soft sensing model for norm vacuum of condensor based on data fusion is proposed due to the enhancement of the accuracy and reliability.

The phenomena of condenser vacuum slowly running down occurred many times in Yinchuan Congeneration Power Plant.

During the test of the condenser vacuum of the power units in Phase 1 & 2,Beilun Power Plant in the summer,the problems of limited suck capability of the vacuum pump and affected vacuum improvement of the condenser in low loading operation have been found.

Has been sufferine from such problems as low condenser vacuum,too high temperature of the casing at stage 17 ang low effec of # 5 LP heater since it put into operation.

The establishment of multi-dimensional data model for condenser by using algorithm of Fuzzy C-means(FCM) cluster has been presented,establishing a data cube for the vacuum in condenser system,and a multi-dimensional analysis for the data cube of vacuum in condenser being carried out by using on-line analytical processing method.

27 single-extraction were put into operation some defects,such as low vacuum of condensers,excessive standard for the temperature of 17th cylinder extreme poor heating effect of No.

The vacuum of condenser is increased at the same time.

The vacuum of condenser has an important influence about economy and security of steam turbine,and it was affected by many factors which had biggish couple,so it is difficult to confirm the norm vacuum of condenser.