Localization of partial discharges in transformers by the analysis of the acoustic emission

The current work presents a solution for the problem related to Partial Discharges (PD) source localization in oil-isolated transformers, using Genetic Algorithms. As known in literature, the simple PD detection is not enough to take a decision about intervening, so the localization is necessary to assess the risk and to plan corrective actions. The parameters extracted from the acoustic signals collected by piezoelectric sensors installed outside of the transformer, provide the correct position of PD. From parameters extracted of acoustic signals collected from piezoelectric sensors installed outside the transformer, the proposed algorithm provides the position where the PD is occurring. This work compares the performance of the method using Genetic Algorithms and those using iterative method, which is normally used to solve this problem. The conclusion describes the advantages of the proposed solution.