Statistical Similarity Based Identification of Power Transformer Inrush Current

Yunfei Luan,Hao Li,Ruochen Zeng,Chun Mo,Luliang Zhang,Mengshi Li
DOI: https://doi.org/10.1109/acpee53904.2022.9783622
2022-01-01
Abstract:Correctly distinguishing internal fault current from magnetization inrush current is the key to differential protection of power transformer. The statistical characteristics of typical disturbance currents in transformer are analyzed in this paper, and propose an identification method for inrush current based on statistical similarity. Ideally, differential current in case of internal fault of transformer approximates sine wave while inrush current approximates half-sine wave. A comprehensive criterion based on mean, variance, skewness and kurtosis is constructed and identify disturbance types by comparing differential current of transformer statistical similarity to standard full-sine wave or standard half-sine wave. When the similarity difference is not obvious, the auxiliary criterion is added to improve the accuracy of identification for inrush current. Simulation results by using PSCAD show that fault current and inrush current in extreme conditions can be distinguished effectively by the proposed method.
What problem does this paper attempt to address?