Dynamic induction machine models including magnetic saturation and iron losses

Dynamic induction machine models are used as the basis for the design and implementation of control algorithms. Costs can be reduced by applying speed-sensorless control, and advanced control strategies open up for the possibility of using an induction machine in demanding applications. However, a reliable and good control performance requires more detailed induction machine models. This thesis deals with models including the magnetic saturation and iron losses. A small-signal model, which includes the saturation due to variations in the main flux magnitude and the load torque, is used to analyze the transient behavior of the machine. Due to the magnetic saturation, the inductances vary as a function of the operating point, and the machine appears to be salient in transients. Based on the model, an identification method for the leakage inductance is proposed. The identification is based on signal injection and can be performed as the machine is running under different load conditions. A model for the skin effect of the rotor bars can be used in combination with the leakage inductance identification in the case of an induction machine equipped with deep rotor bars. The magnetizing curve can be modeled using a simple power function. An adaptive identification method is developed for the identification of magnetizing curve parameters. Identification of the leakage inductance prior to the magnetizing curve identification improves the results in case a no-load condition cannot be reached. The stator hysteresis and eddy current losses are modeled using a nonlinear resistance. The resistance is not dependent on any frequency, and is thus defined also during transients. The resistance model is experimentally investigated both for the case of an induction machine and a nonlinear inductor. The iron loss model is used in a loss-minimizing control algorithm for the induction machine.