Permeance Network Based Real-Time Induction Machine Model

In this paper a real-time permeance network based distributed model for induction machines is proposed which can accommodate the local phenomena such as slotting and space harmonics inside a machine. A linear permeance network model (PNM) of a squirrel encage induction motor is successfully simulated in real-time by optimizing the numerical procedures inside the model. Particular attention is paid to the selection of an efficient linear solver for the model. Also, the moving band technique which is already used in the finite element analysis of electrical machines is extended to the PNM to reduce the domain of study to one quarter of the machine. Real-time oscilloscope results of the simulation of an induction machine during the start up, no load and loaded operating conditions are shown in the paper. The real-time simulation results are validated by comparing the effect of rotor slot harmonics in the frequency spectrum of the steady-state stator current with analytical formulas. In order to investigate the effect of saturation on the performance of the machine, a comparison between the simulation results from the real-time linear PNM and offline
nonlinear PNM is also presented in the paper.