Generalized High-Fidelity Reduced-Order Modeling of Doubly-Fed Machines and Induction Machines

Abstract

Doubly-fed induction machines (DFIMs) have been widely used in wind power generation and high power drives. The brushless versions further improve the reliability and reduce the maintenance cost by eliminating brushes and slip rings, showing great potential in replacing the traditional DFIMs. The successful model-based control development of doubly-fed machines relies heavily on the high fidelity of the plant model, which most of the existing DFIM models lack. This paper proposes a generalized high-fidelity modeling approach by extracting lookup tables from parametric finite element analysis (FEA) sweeps for brushed and brushless doubly-fed machines. It can be applied directly to induction machines by shorting the rotor circuits. The developed model covers the magnetic saturation of machine cores and spatial harmonics introduced by the winding layout and slotting effect. It is suitable for system-level simulations and controller development of doubly-fed machines and induction machines, either in motoring or generating modes, with pure sinusoidal or pulse-width-modulation (PWM) voltages. Detailed modeling implementations and simulation results of a slip-ring DFIM are provided to validate the modeling workflow and showcase the fidelity level compared to time-consuming electromagnetic transient FEA.