Parameter identification of power transformers thermal model via genetic algorithms

Recent studies by various authors have shown as the IEEE Transformer Loading Guide model and the more recent modified equations, proposed by the Working Group K3 of the IEEE ‘Power System Relaying Committee’, are lacking in accuracy in predicting the winding hottest-spot temperature of a power transformer in presence of overload conditions. This is mainly due to the deviation of the parameters of the thermal model of the power transformer in the presence of overload conditions. In the paper, a novel technique to identify the thermal parameters to be used for the estimation of the hot-spot temperature is presented. The proposed method is based on a genetic algorithm (GA) which, working on the load current and on the measured hot-spot temperature pattern, permits to identify a corrected set of parameters for the thermal model of the power transformer. Thanks to data obtained from the experimental tests, the GA based method is tested to evaluate the performance of the proposed method in terms of accuracy