Online monitoring of transformer winding axial displacement and its extent using scattering parameters and k-nearest neighbour method

The online monitoring of the transformer winding using scattering parameters fingerprint is presented. As a test object, a simplified model of transformer is used. The winding axial displacement is modelled on this test object. The scattering parameters of the test object are calculated using the high-frequency simulation software and measured using a network analyser. Two indices are defined based on the magnitude and phase of scattering parameters for the detection of the axial displacement. A new algorithm for the estimation of the axial displacement extent is presented using the proposed indices and high-frequency modelling of the transformer. To detect this mechanical defect and its extent, the k-nearest neighbour (k-NN) regression is suggested.