Study On Frequency Response Analysis Method In Diagnosing Transformer Winding Deformation

Transformer is one of the important equipments in the power system, and its running state influences the system's security directly. The transformer can be collided when being transported or installed, and shorted abruptly when running, which can lead to the winding deformed. The winding deformation is one of the familiar faults of the transformer, and it harms the power system's secure operation. For the purpose of detecting the latent accidents in time, it becomes very important t to diagnose the winding deformation.As one of the methods of diagnosing the winding deformation, the frequency response analysis method is presently being widely used in many countries besides China. By far there are no settled scales or uniform criteria that can be abided by to diagnose winding deformation, while its use is mainly depended on the comparison of the frequency response curves. Therefore, the development of diagnosing technique lags behind the measuring and testing techniques. The main aim of this dissertation is to study how to detect and diagnosis transformer winding deformation on the base of frequency response curves.In the first chapter of the dissertation, the methods and their merits and defects of diagnosing the winding deformation are summarized. Subsequently, the development of frequency response analysis method is presented.In the second chapter, the structure of transformer is described, and the method of calculating the short current and the distribution of oozy magnetic field of winding are put forward.The tests analysis are presented in the following two chapters, and the relation between the frequency response curves and the change of parameters is to be obtained by means of altering parameters of different site's inductance and capacitance. In the third chapter, the equivalent model of the winding is simulated using PSPICE. The tests of the practical model are given in the fourth chapter. The testing system, connection mode and influence facts in testing are discussed, then the calculation of winding's parameters is presented. Considering the influence facts in the tests, the mathematic model of the transformer is established using the parameters and the frequency response curves.In the fifth chapter, the theory of using orthogonal polynomials to estimate the transfer function based on the least square methodology is introduced, then the zeros and poles are obtained from the transfer function. The detection method by using the poles distribution is better than the method only depending on frequency response curves, because the poles and zeros of the transfer function can reflect the essence of the windings. In the last part of thischapter, the effects of the winding deformation imposed on the distribution of poles and zeros are analyzed integrating the actual transformer. In the last chapter, the whole text is summarized and some questions to be perfected are brought forward.