Study on Evolution Characteristics of Partial Discharge in Transformer Insulating Oil

As the core equipment of AC power grids, the oil-paper insulation structure of power transformers is prone to accidents such as explosions and fires due to frequent partial discharges. In order to suppress the development of transformers to severe arc discharges and explore the dynamic evolution characteristics of partial discharges from the source, this paper builds an experimental platform for partial discharge image acquisition in oil from the perspective of morphological analysis. Based on a multi-scale retinal algorithm with grey-scale analysis, an enhancement processing algorithm for low-illumination discharge images was designed to extract a continuous discharge profile. The effect of the external applied voltage on the horizontal displacement and half-perimeter distortion of the discharge was then explored. The results show that the horizontal displacement of the discharge at low voltage levels is small and the aberrations are not obvious. At high voltage levels, the discharge horizontal displacement and morphological distortion jumped strongly, and the energy was not easily concentrated. The discharge in the transition stage is always maintained at a high energy level, instead showing the least number of discharges.