Dynamic Modal‐Based Vibration Characteristics Analysis and Looseness Fault Diagnosis in Transformer Winding

ABSTRACT Limited by the theoretical basis and diagnostic methods, the winding mechanical condition monitoring often requires both normal condition vibration signals and present vibration signals, and the winding mechanical condition identification is completed by comparing with the normal condition signals. This paper proposes a method to diagnose winding looseness faults without the need for normal condition vibration data based on the established dynamic modal model (DMM). Starting from the vibration mechanism, the DMM of the winding is established, and the mutual coupling relationship between the dynamic modal (DM) and vibration is clarified. Then, a numerical method is used to simulate the variation pattern of the winding natural frequency under the influence of the DM; an experimental method is used to study the difference between the frequency response curve under the influence of the DM and the frequency response curve of the static modal (SM). According to the above research results, the phenomenon of vibration deviation under the same mechanical condition and the identical excitation is revealed. Finally, taking into full consideration the characteristics of vibration deviation and changes in vibration itself under the effect of the DM, a diagnostic method for identifying winding looseness faults without the need for normal condition signals is proposed. The research results provide a practical solution for identifying the mechanical condition of the winding in the absence of normal condition data.