Research On Simulation And New Protection Algorithms Of Transfomer

With the development of the power system especially with the evolution of the modem new material and new arts, and the capability of transformer becoming bigger and bigger, the requirement for improving performance of transformer protection relay becomes stronger. As to study the protection relay, the first thing needs to do is to analyze the mechanism and the feature of faults inside transformer, then raises a reasonable protection scheme. The target is to exploit the relay with high performance. The transient simulation of the fault transformer is one of the indispensable qualifications for improving the performance of transformer protection relay. The dissertation includes the author's research on protection relay and simulation of transformer in recent years, and also introduces some new achievements in this field. The key technique of the transformer's transient simulation is the treatment of ferromagnetic elements loop. A new Hybrid Radial Basis Function Networks used to simulate the ferromagnetic elements major hysteresis loop is proposed in this dissertation, and the fractal phenomena of the ferromagnetic elements hysteresis loop is firstly brought forward. The mutuality of ferromagnetic elements major hysteresis loop and its hypo-loops are revealed using the fractal theory. According to iterated function system of fractal theory, compressing the major ioop can form the hypo-loops. The result of the simulation has validated this idea. The internal fault of transformer's windings takes place frequently. The simulation of transformer's internal fault is an important tool for the study of transformer protection relay. A method of modeling internal faults of power transformer is taken with in this dissertation, the model is entirely compatible with the EMTP software, it enables any two turns of the transformer windings, the validation of the method is right by the simulation. Inrush current identification is a key problem of transformer protection. In this dissertation, a novel theory based on the Auto Correlation Analysis Method is firstly offered to distinguish between inrush current and fault current. In this dissertation, many characteristics of inrush current is used to distinguish between inrush current and fault current in the Auto Correlation Analysis Method. The Simulation studies with respect to different fault and inrush conditions have been conducted, and the results prove that the proposed technique is able to offer fast responses in protection and accurately discriminate between inrush current and internal faults. A fuzzy mathematics origin from modem mathematics, the theory of fuzzy close measure is a powerful tool of pattern identification. Fuzzy close measure theory is firstly introduced to the inrush identification, a new algorithm based on forecasted waveform is brought forward in this dissertation. The standard sine waveform is forecasted from fault current or inrush current with a short sampling data window, analysis of practical sampling waveform and standard sine waveform can be correctly identified inrush current form internal fault current using fuzzy close measure. This method is proved to be highly effective and fast by the simulation. It is an effective approach to improve the performance of transformer protection relay by introducing new theory and new technique to transformer protection relay. In this dissertation, a new algorithm based on wavelet theory is brought to distinguish between inrush current and fault current. The tran...