Modeling and Emulation of Induction Machines for Renewable Energy Systems

Electric motors with their drive systems utilize most of the generated power. They account for about two-thirds of industrial energy utilization and about 45% of the global energy utilization. Therefore, optimizing a motor and its corresponding drive system with their control can save energy and improve system efficiency. It can be risky and difficult however to test large prototype electrical machines and study their dynamic characteristics; or to test electric drives at high power levels with different machine ratings and operating conditions. One way to effectively evaluate these systems is to emulate the electric machine using a power electronic converter with the help of a real-time simulation system. Power electronic converters and their control systems are increasingly being used in industries with different power ratings at high switching frequencies in the kHz range. In this thesis, an induction motor emulator based on a power electronic converter is developed to allow detailed testing the converter and controller. A proportional-resonant current controller in the abc-frame and pulse-width modulation are employed. The conventional model of the induction machine (IM) with constant parameters does not represent accurately the machine’s performance for severe transients specifically during starting and loading conditions. Magnetic saturation effects should be considered. Hence, experimental procedures to determine the flux saturation characteristics in the main and both stator and rotor leakage flux paths are achieved. Machine models that consider or neglect the main and leakage flux saturation are compared with experimental results. The model which considers the magnetic saturation effect in both flux paths results in more accurate transient responses. Likewise, the dynamic response of the induction motor emulator during startup and loading transients show the effectiveness of using the developed emulator to resemble closely a real motor. The relationship between the stator and rotor leakage reactance of the induction machine according to IEEE Std. 112™ is assumed to be constant under all operating conditions. However, this is not accurate during severe transients such as the direct online startup and loading conditions of a three-phase induction motor. The leakage reactance of the machine can vary widely during severe conditions. Hence, using constant parameters in the machine model will result in an inaccurate dynamic performance prediction. Moreover, considering a constant ratio between the stator and rotor leakage reactance is no longer valid for all current levels. In this thesis, a direct and precise method is proposed to estimate and separate the stator and rotor leakage reactance parameters under normal operating conditions and when the core is deeply saturated. The method exploits the 2D time-stepping finite element method (FEM) with a coupled circuit. The obtained current-dependent reactance functions in both leakage flux paths are included in the dq-model of the IM. Other machine parameters are determined by implementing the standard tests in FEM. To verify the effectiveness of the proposed method, the predicted results are compared to the dynamic responses obtained experimentally from a three-phase, 5-hp squirrel-cage IM. A power electronic converter-based self-excited induction generator (SEIG) emulator is developed. The testbed replaces a wind- or microhydro-turbine driven squirrel-cage induction generator that works within an isolated power system to feed power in remote areas. It supports testing and analyzing the dynamic performance of islanded generation systems which comprise numerous kinds of parallel-operated renewable energy sources. The risk and cost associated with the testing, analysis and development of novel control topologies and electrical machine prototypes are reduced considerably. The dq-model of the SEIG in the rotor reference frame is implemented in a real-time controller. Saturation in the main and leakage flux paths are included in the machine model. The generator model with modified parameters is verified and used in the emulator. The cascaded voltage and current loops utilizing the proportional-integral controllers in dq-frame are employed. A voltage-type ideal transformer model is used as a power interface for the emulator whereas an excitation capacitance is added to the power-hardware-in-the-loop block diagram. Likewise, the dynamic response of the induction generator emulator during voltage buildup and loading conditions validates the effectiveness of using the developed emulator to resemble closely a real generator