Estimation of the parameters of metal oxide gapless surge arrester equivalent circuit models using genetic algorithm

In the present work a genetic algorithm is developed, for the evaluation of the parameters of metal oxide gapless surge arrester circuit models, in order to minimize the error between the computed and the measured (by the manufacturer) peak value of the residual voltage for each given current waveform and level separately. Furthermore, the algorithm is modified in order to minimize the error simultaneously for all the given injected impulse (lightning and switching) current curves.

► We estimate the circuit parameter of surge arresters using genetic algorithm.
► Manufacturers’ data are used for the estimation of the parameter values.
► We examine the influence of three different models in the accuracy of our results.
► We have accurate results for wide range of injected currents (lightning & switching).