Permeance network modeling of the stator winding faults in electrical machines

The aim of this paper is to present a robust model based on permeance network method (PNM) for the diagnosis of electrical machine stator faults. The proposed model allows taking into account the local magnetic saturation of the magnetic circuit due to heavy stator fault currents and remains moderately time consuming compared to the finite elements method. Firstly, the system of algebraic equations describing the developed permeance network is written. Then, this magnetic model is coupled to the electrical differential equations describing the induction machine in presence of inter-turn stator faults. Finally, simulation results are successfully compared to the experimental results. The impact of saturation on inter-turn fault signature clearly confirms the interest of a nonlinear model based diagnostic for this fault. Index Terms--Electric machines and drives, modeling, simulation, permeance network, diagnostic.