Research On Identification Of Transformer Inrush Currents Based On Virtual Instrument Technology

As the vital importance to power transmission link, the transformer’s normal operation relates to the safety of the power system, and it’s difficult to measure the economic losses and disaster consequences of the transformer failure because of the large power transformer’s cost. As the key issues in the transformer differential protection, the recognition accuracy of inrush current relates to the right action rate of the differential protection, and to the power quality. So, many scholars and engineers domestic and overseas made extensive exploration and put forward a variety of identification methods which have advantages and disadvantages. The methods cannot fully meet the requirements of relay protection, as the voltage grade and transformer capacity were greatly improved. This is why looking for a new type of inrush current identification method still has urgency reality significance.In this paper, based on the research and analysis on the formation mechanism of the transformer inrush current and fault current, they are both mathematically simulated in the Lab VIEW software, and then the waveform distribution coefficient histogram is used to analyze the differences. And the optimal threshold is calculated by the Otsu method which is widely used in digital image threshold segmentation, and due to the different characteristics of peak valley, the best threshold values are different. Then, inrush current identification method based on Otsu method is established and the mathematically simulated, under the condition of simplified core magnetization curve, the identification result is correct. The data acquisition function of virtual instrument, with the modulation circuit, combines mathematical simulation and physical simulation, and as a result, the semi-physical simulation, also called as the Hardware-in-The-Loop simulation, which is summed their advantages is realized in the platform. The actual transformer inrush current is sent into the program to experiment in this simulation platform. A large number of no-load switching-in experimental results show that this identification method has accuracy and reliability.