Title :
Signal trend identification with fuzzy methods
Author :
Wang, Xin ; Wei, Thomas Y C ; Reifman, Jaques ; Tsoukalas, Lefteri H.
Author_Institution :
Sch. of Nucl. Eng., Purdue Univ., West Lafayette, IN, USA
Abstract :
A fuzzy logic-based methodology for online signal trend identification is introduced. Although signal trend identification is complicated by the presence of noise, fuzzy logic can help capture important features of online signals and classify incoming power plant signals into increasing, decreasing and steady-state trend categories. In order to verify the methodology, a code named PROTREN is developed and tested using plant data. The results indicate that the code is capable of detecting transients accurately, identifying trends reliably, and not misinterpreting a steady-state signal as a transient one
Keywords :
fuzzy logic; noise; power plants; signal classification; PROTREN; fuzzy logic; noise; online signal trend identification; power plant signal classification; steady-state signal; transient detection; transient signal; Data mining; Fuzzy logic; Inductors; Laboratories; Power generation; Power system reliability; Signal processing; Steady-state; Testing; Thermal management;
Conference_Titel :
Tools with Artificial Intelligence, 1999. Proceedings. 11th IEEE International Conference on
Conference_Location :
Chicago, IL
Print_ISBN :
0-7695-0456-6
DOI :
10.1109/TAI.1999.809813