PCA based protection algorithm for transformer internal faults

Authors : Erdal KILIÇ, Okan ÖZGÖNENEL, Ömer USTA, Dave THOMAS

Pages : 125-142

View : 14 | Download : 8

Publication Date : 0000-00-00

Article Type : Research Paper

Abstract :In this paper a new protection scheme is introduced to detect and identify transformer winding faults. The new approach is based on artificial neural networks insert ignore into journalissuearticles values(ANNs); using radial basis functions insert ignore into journalissuearticles values(RBFs); and the principal component analysis insert ignore into journalissuearticles values(PCA);. The nonlinear system`s input and output data is manipulated without considering any model of the system. This approach is used to detect and identify internal short circuit faults of a three phase custom built transformer. The suggested technique is also able to distinguish between the fault and magnetizing inrush current. The test studies carried out shows that the proposed method leads to satisfactory results in terms of detecting and isolating parameter faults taking place in non-linear dynamical systems.
Keywords : Protection, internal faults, transformers, PCA, ANN, RBF