Research On Digital Simulation And Novel Digital Main Protective Principles Of Transformer

Transformer is one of major electric apparatuses in power system. It plays an important role in the transmission of power supply. If it had a fault, the security and stability of power system would be greatly affected. Modern EHV transformers with large capacities have evolved to meet the increasing need of power system. On the other hand, if EHV transformers were damaged by faults, it would need a long time and great effort to repair them, for which, great economic benefits would be lost. Considering the serious consequence caused by faults in transformer, the research for new algorithm of transformer protection with high performance, such as fast, reliable and sensible capabilities, is in highly demand. The dissertation makes an investigation on some key problems, which are not well solved and still exist in transformer protection. Based on deep analysis and research, new algorithms to solve these problems are provided. The major contributions of this dissertation are as follows: 1)The dissertation proves that ferromagnetic element's hysteresis loops is a kind of the fractal figure using the basic conception and definition of the fractal theory. Rules obeyed in the dynamic magnetizing procedure of ferromagnetic element are raised. A new algorithm based on the fractal theory that uses the "variable compressing factor" on simulating minor magnetic loops is put forward. Minor magnetic loops are emerged by using a linear-varied factor in compressing the major magnetic loop. Simulation results validate this idea. 2)A new algorithm of distinguishing the inrush and fault current of power transformer based on the character of current waveform extracted by morphological gradient is creatively put forward. Morphological detector, which is used to locate the peak and the bottom of current waveform, is designed. By calculating the morphological gradient between the peak and the bottom point of waveform, "three-area-compare" algorithm is originally put forward, which can be used to make the qualitative discrimination between the inrush and fault current. Protective setting values and parameters of transformer are no more needed. The proposed algorithm is able to distinguish the inrush and fault current with good performances, such as immune to CT's saturation, little affected by decaying DC component and few calculations .etc. 3)The dissertation makes a deep analysis on the calculation of the equivalent instantaneous inductance of wye-connected transformer. It concludes that the looping current in the winding of delta side of transformer, which is hard to obtain from line current, severely influences the calculation result of equivalent instantaneous inductance. In that case, a new scheme of transformer differential protection is proposed. Considering the variation of the equivalent instantaneous inductance, an new algorithm based on the normalized equivalent instantaneous inductance for the discrimination between inrush and fault current is creatively proposed. 4)To prevent the mal-operation of differential protection because of current transformer(CT) saturation at out-zone fault, an algorithm based on the mathematical morphology is first put forward in this paper. Unlike the traditional time-differential method, this algorithm only needs detecting the time of fault occurred, while the time of differential current increased is no more needed. In the case of CT's serious saturation, a discrimination between out-zone fault and inner-zone fault can be clearly made by this algorithm, which needs a few calculations and a short data window. In addition, the proposed algorithm can also be applied to bus protection. 5)A new transformer protective scheme based on the improved sequence-impedance algorithm is put forward in this dissertation. This algorithm has good performance in distinguishing transformer's internal and external faults, and is immune to CT's saturation and ratio-mismatch at certain extend. Based on the achievements of formers, this dissertation contributes in the following aspects, such as improving...