91 / 2021-06-05 12:28:33

PROPAGATION AND IDENTIFICATION OF PARTIAL DISCHARGE SIGNALS IN TRANSFORMER UNDER THE FRAMWORK OF DIGITAL TWINS

The safe and stable operation of the transformer is of great significance for maintaining the safety of the power grid and improving the stability of the power supply. Most of studies have shown that the main reason for transformer failures is the insulation deterioration caused by long-term internal partial discharge (PD) under operating conditions [1]. During the propagation, the discharge signal may be attenuated by the refraction and reflection of the discontinuous structure. The signal acquired at the detection end tends to be distorted, which increases the difficulty to identify the discharge signals [2]. In recent years, with the development of digital twin technology, it has become easier to study the characteristics of the signal propagation in transformers. As shown in Figure 1, this paper first acquired the signals of three typical defects, which are protrusion, surface discharge, and gas gap in the experimental transformer. Then, they were applied as a signal source to the distributed parameter model of the transformer winding shown in Figure 2. The corresponding detection signals and the propagation features of the signals of different detect types were acquired. By extracting features such as rising time, falling time, signal amplitude, and steepness of the detected signal, and using linear discriminant analysis algorithm (LDA) to reduce dimension, we acquired a good signal recognition result. This shows that even after the attenuation of the propagation path, the detected discharge signals still contain the information related to the detect type, which can be used to identify the discharge signals.