Air-gap eccentricity fault detection, isolation, and estimation for synchronous generators based on eigenvalues analysis

Abstract

Air-gap eccentricity fault causes rotor-stator rub and consequently damage to Synchronous Generators (SGs). In this paper, a fault diagnosis approach to diagnose the eccentricity fault for SGs is presented. In this approach, the state matrix eigenvalues based on the subspace identification are estimated, and those are used for fault diagnosis. Two dq models of SGs in faulty and healthy conditions are employed to present the theoretical foundation of the method. As the main advantage, the introduced fault diagnosis method is working properly for either linear or nonlinear loads of SGs. The stator and field currents and voltages, and rotor rotational speed are required signals in the introduced approach. The method is validated using experimental data of SGs in Iran-Safir (ER24) diesel-electric locomotives.

Keywords: Air-gap eccentricity; Fault diagnosis; Subspace identification; Synchronous generator.