A New Modified Hyperbolic S Transform-Based Method for Discrimination of Internal Fault from Other Abnormal Conditions in Power Transformer

In this paper, a new method is presented for discrimination of internal fault other abnormal conditions in power transformers. Since the differential current signals are non-stationary, a powerful tool such as S transform is needed for time- frequency analysis of them. First, by considering the effective factors, different simulations for internal fault, inrush current, external fault, over-excitation and ultrasaturation conditions are performed via PSCAD/EMTDC software in different operating conditions of transformer.Then, the obtained differential current signals are analyzed by modified hyperbolic S transform. Important characteristics are extracted from modified hyperbolic S transform and according to them, different conditions are distinguished and classified. These features include: energy contour of first level, variance index, standard deviation of the fundamental frequency and second harmonic component and three special criteria. The process of modified hyperbolic S transform and performing the proposed algorithm are performed in MATLAB software. In this paper, all of possible transient conditions and effects of the saturation of current transformers are considered. Finally, the accuracy and precision of the proposed method are compared with different methods in conditions with noise and without noise. The results show good performance of the proposed method.