Title :
Identification of the Nonlinear Model proposed by the MIT for Power Transformer applying Genetic Algorithms
Author :
Pérez, R. ; Matos, E. ; Fernández, S.
Author_Institution :
Dept. de Ing. Electr., UNEXPO Vicerectorado de Barquisimeto, Barquisimeto, Venezuela
Abstract :
This paper presents a technique based on genetic algorithms for the parameter estimation of the top oil temperature nonlinear model in power transformers proposed by MIT used in on line diagnosis and monitoring systems, installed in a 100 MVA 230/115/24 kV OA/FA/FOA power transformer of Substation Barquisimeto ENELBAR Venezuela since 2003. The results of the parameter estimation by genetic algorithms are compared with parameter estimation by least-squares and measured top oil temperature. These results are discussed and the authors proposed this model as power transformer diagnosis valuable tool.
Keywords :
genetic algorithms; least squares approximations; parameter estimation; power transformers; MIT; Substation Barquisimeto ENELBAR Venezuela; apparent power 100 MVA; genetic algorithms; least-squares; line diagnosis; monitoring systems; nonlinear model identification; parameter estimation; power transformers; voltage 115 kV; voltage 230 kV; voltage 24 kV; Condition monitoring; Genetic algorithms; Parameter estimation; Petroleum; Power system modeling; Power transformers; Silicon compounds; Substations; Temperature measurement; diagnosis; genetic algorithms; parameters estimation; power transformers;
Journal_Title :
Latin America Transactions, IEEE (Revista IEEE America Latina)
DOI :
10.1109/TLA.2009.5419360
