Application of ANN and Equivalent Electrical Circuit for Understanding the Power Transformer Frequency Response Analysis

Frequency Response Analysis (FRA) is a well-known method approved and increases interest in performing electric power utility tests. FRA has become a recognized technology for detecting winding and core deformations in power transformers. The weakness of the FRA method is that there is no approved standard or guidelines to understand the measured results, which depends on individual knowledge. For further understanding of the FRA measurement, there are the lumped circuit approaches for power transformer fault analysis. The presented circuit approaches elements of the transformer, including windings and core, can be measured, or calculated. The challenge of identifying parameters for a power transformer equivalent circuit can be solved using frequency response analysis (FRA) and artificial neural networks (ANN). In this paper, the two techniques proposed and employed to extract the appropriate values of transformer circuit parameters that can simulate the identical measured response. The proposed technique is examined through its application on a three-phase transformer with 33/11 kV 500kV A. The FRA measurement was conducted using the end-to-end open-circuit test. The results show that FRA and ANN techniques can be used to identify transformer-equivalent circuit parameters. Furthermore, the proposed methods can be extended and used to estimate the parameters of a three-phase power transformer of different ratings.