Transient Processes And Optimization Of Winding Insulation For Power Transformer

The transient process of winding insulation for power transformer is a potential threat to the power transformer insultion. The violent transient process may cause damage to the transformer insulation, or even it may lead to many serious accidents, such as the explosion of transformer and the failure of powr system. Therefore, the analysis on the transient process of winding insulation and its optimization for power transformer is particularly important.Based on the equivalent inductance parameters, capacitance parameters and resistance parameters of power transformer, the equivalent circuits for initial potential distribution and for oscillatory potential distribution were built, respectively. The influence of the total continuous windings, the intershield-continuous windings and the interleaved-continuous windings on the transient process of winding insulation was analyzed. The initial potial distribution, the steady potential distribution and the osillatory potential distribution of the intershield and continuous windings for power transformer was discussed when the neutral point of power transformer was grouded or insulated. At last, on the basis of the potential gradient of winding insulation, the genetic algorithm is used to analyze the optimization performance of winding insulation for the power transformer, and the simulation verifies the optimization effect.The results show that the intershield-continuous or the interleaved-continuous winding can obviously improve the intial potential distribution and the potential gradient of the power transformer. The intial potential distribution of the power transformer is hardly affected when the neutral point is grouded or insulated. The greater difference between the intial potential distribution and the steady potential distribution, the more violent oscillatory process appears to occure. The genetic algorithm can enable the optimization of the winding insulation for the power transformer.