Thermal model parameters identification of power transformer using nature-inspired optimization algorithms

Abstract

The thermal property of transformer insulation is influenced by the design and size of transformer concern. As non-interrupted power supply is required to avoid financial loss and to provide reliable service to the consumer, thermal modeling of such power supply equipment has become an important tool for condition monitoring of power transformer. Thermal model equations are used to determine the hot spot temperature as it represents thermal condition and heat dissipation of transformer. This paper proposes a methodology to identify the thermal model of oil-immersed power transformer insulation. In this paper, measured top-oil and bottom-oil temperatures of a power transformer are used to find parameter of simple thermoelectric analogous thermal model. Nature-inspired algorithms are used to parameterize the thermal model. Three algorithms, namely genetic algorithm (GA), particle swarm optimization (PSO), and simulated annealing (SA), are used for the purpose. The cost functions of these algorithms are based on thermal model equations that are reported for the prediction of top-oil and bottom-oil temperatures. In addition, this paper also presents a comparative analysis between these optimization techniques.