Partial Discharge Identification of Power Transformers Based on Chaotic Characteristics of the Cumulative Energy Function

The diagnosis of partial discharge(PD) in power transformer is of great significance to the evaluation of transformer operation state. The traditional method which is based on pattern recognition of Phase Resolved Partial Discharge(PRPD) Pattern is limited to the morphological characteristics of partial discharge spectra analysis. The lack of deeper research of the characteristic of partial discharge leads to low recognition rate and inaccuracy. This paper presents a method of partial discharge pattern recognition based on the chaos theory and the cumulative energy function. A chaotic analysis of the cumulative energy function of partial discharge signals is carried out to calculate the maximum Lyapunov exponent and obtain the maximum Lyapunov exponent distribution characteristics of different types of partial discharge defects. In this paper, different types of partial discharge defects, such as protrusion discharge, void discharge and surface discharge, were placed on the 35kV experimental transformer. The experimental results show that the overall recognition of partial discharge type of transformer is better than traditional method by using chaos characteristics.