A Robust Approach for Locating and Assessing Mechanical Faults in an Actual Transformer Winding Using the State Space of Its Lumped Equivalent Model

Abstract

Abstract—Transfer function and state-space model are two powerful complementary tools have been used to monitoring the physical system. These two ways can be very successful in giving a comprehensive analysis of electrical power transformer malfunctions. This paper proposes a novel robust approach nondestructive to diagnose short circuit (SC) faults by considering axial displacements (AD) in the transformer winding. Initially, elaboration a native autonomously tool to control the LCR meters through computing systems. This platform will control the impedance measurement by the instrument of the LCR meter through MATLAB software. Additionally, achieving estimation autonomously to all the parameters of the lumped equivalent model of the winding accurately using frequency response analysis signal morphology interpretation (FRASMI). On the other hand, obtaining a new representation of the state-space model of the lumped equivalent model of an isolated winding with SC fault between its turns to locate and assess its severity. Finally, the relationship between the model parameters and the resonance frequencies of the frequency response curve can provide an effective nondestructive diagnostic tool for all mechanical damages of winding (e.g., SC with AD) which have been implemented experimentally, in which a new physical index has been proposed making it possible to locate and assess the extent of these defects simultaneously