Study on chaotic characteristics of time series of partial discharge in oil-paper insulation

Abstract

In the past, partial discharge was considered as a completely random process, and various statistical spectra such as phase resolved partial discharge (PRPD) were constructed, but the accuracy still needs to be further improved. The chaos theory is introduced to research the development of partial-discharge defects of oil-paper insulation in transformer. The sphere-plate partial discharge model was constructed in the laboratory. And the data of the pulse discharge during the development of the pressboard from the initial discharge to the breakdown was obtained by the constant voltage method under six voltage levels. By constructing the discharge time series, the chaotic attractors are reconstructed in the phase space. And all the attractors of the samples under the six voltage levels satisfy the largest Lyapunov exponent λ 1 >0, the correlation dimension D is the fraction and the Kolmogorov entropy K 2 >0, indicating that the partial discharge is a deterministic chaotic process. It provides new research ideas for the diagnosis and prediction of the aging and the development of electrical branches for oil-paper insulation.