Study of the Relation Between EnvironmentalConditions and Insulators leakage Current in aDistribution Network Using Artificial NeuralNetworks

The leakage current which flows throughinsulators is dependent on many different environmentalvariables like temperature, humidity, wind and pollution. Theleakage current rate varies during a daily period; so the leakagecurrent curve versus time in a ۲۴ hour period will be nonlinear.Implementing field experiments in Hormozgan Research Site,the leakage current peak amounts were collected in differentintervals and using two neural network methods of FeedForward BP algorithm and Radial Based Function, the relationbetween leakage current and variation of effective parameters ofleakage current was studied and the most effective algorithmrevealed.