A design of inrush current identification system for high-speed train´s traction transformer

Distinguishing inrush current and short circuit current is a key issue for the longitudinal differential protection of high-speed train´s traction transformer. This paper introduced virtual instrument technology to build an inrush current identification system, which utilized software to realize the identification algorithm and applied digital signal processing approach to the differential protection of the transformer. It developed the system software in virtual instrument tool LabWindows/CVI. The algorithm regards the inrush current and short circuit current as random signals, calculates the self-correlation functions of sampling data, then compares them with standard sine self-correlation coefficient vector table, and finally uses the magnitude of similarity coefficient to distinguish inrush current and short circuit current. Theoretical analysis and experiment result have proved that the system has high accuracy, fast decision rate and good expansibility.