Research on fault sample knowledge abstraction and fault fuzzy diagnosis for large power station halving-style condenser

Abstract

Typical fault sample knowledge library of the halving style condenser is abstracted and completed based on the dynamic simulation model of halving-style condenser developed by the authors. The fuzzy pattern recognition is used to realize halving-style condenser fault diagnosis. A new kind of subordinate function is given to overcome the demerit of one old function. Diagnostic effects of the two different functions are compared by example.