Modification of IEEE model for metal oxide arresters against transient impulses using genetic algorithms

Dynamic characteristics of zinc oxide arresters are of great importance in studies related to steep front impulse currents. Several models are suggested up to now for frequency analysis simulations with acceptable validity. In this paper, different electrical models which are represented for analysis of transient states, are investigated and then using simulations and comparison between experimental and simulation wave forms, a model is tried to obtain which is more similar to the reality in quantity (number of elements) and in quality (wave form). Within existing models, IEEE model is very suitable in quantity and quality but it looses its efficiency because of absence of residue voltage caused by switching current in most producers' catalogue (the test is very complex) and therefore troubles modeling of this type. In this paper, a special genetic algorithm is represented to correct IEEE model. Considering that the mentioned method represents very acceptable results in the range of current surge wave front time from 0.5
mu
sec to 45
mu
sec, it is possible to reach a model with appropriate parameter values using a standard impulse (8/20
mu
sec) and one other arbitrary wave with try and error method and finally by stating correcting method for IEEE model, related results are investigated and compared with other models.