A Winding State Evaluation Method Based on the Vibration Correlation Coefficient

Implementing feature mining of transformer winding vibration signals is of great significance for quickly evaluating the operating status of windings. This paper first theoretically analyses the generation process of the the quasi-steady state mode of winding vibration, and based on this, proposes a method to evaluate the winding state using the vibration correlation coefficient. This method believes that when the Pearson correlation coefficient of adjacent signals is higher, the signal stability is better, and the winding structure is more stable. Finally, this method is applied to the state evaluation of an abnormal high impedance, The results validate that this method can effectively distinguish different operating states of windings and provide a new reference for evaluating the state of transformer windings.