Identification and Inversion of the Non-Linear Magnetic Model of Anisotropic Synchronous Machines

This paper proposes a magnetic model inversion technique based on the analytical equations of synchronous machines. The magnetic model of anisotropic machines, such as an interior permanent magnet synchronous motor or a synchronous reluctance machine, is highly nonlinear due to saturation and cross-saturation effects in the orthogonal d, q axes equations. The identification of these models is a well-established subject, but once identified, it is computationally hard to invert those models for simulation and analysis purposes. The inversion problem is tackled here in a rather simple way by exploiting the linear behaviour of the core under highly saturated conditions. The tests are carried out on a commercially available synchronous reluctance machine to identify two magnetic models that are inverse of each other. The analytical inversion method is then used to show that the experimentally identified inverse model indeed can be obtained without computationally involved algorithms.