Identificación y validación del modelo no lineal de la temperatura del aceite superior de transformadores de potencia aplicando algoritmos genéticos

this paper presents a technique based on genetic algorithms for the identification and validation of the top oil temperature no lineal model in power transformers used in on line diagnosis and monitoring systems, installed in a 100 mva 230/115/24 kv oa/fa/foa transformer of substation barquisimeto enelbar venezuela since 2003. the results of the identification by genetic algorithms are compared with identification by least-squares and measured top oil temperature. the results demonstrate a significant reduction of the mistake in the model when making the identification by genetic algorithms, which improves its performance like as power transformer diagnosis tool.