Abstract

Recent engineering applications such as electric and hybrid electric vehicles require higher performance electric machines. Accordingly, there have been a significant increase on research on performance improvement of electric machines for transportation applications. Opposed to the conventional machines, new and innovative traction machines have higher power density, higher efficiency, faster dynamic response, wider speed range and higher reliability. The modern simulation and manufacturing tools made it possible to obtain the desired performance requirements in the new and special designs at reasonable cost. Due to the special designs to improve a certain output variable subjected to several constraints, it is required to have advanced machine models for control and operation of these machines. Moreover, parameter measurement of these special electric machines are gaining a significant interest due to new and advanced motor drive testing systems such as power hardware in the loop emulation which use a look-up table based machine model for fast and accurate solution of machine dynamics.
This PhD work develops a novel automated current control method to measure the parameters of special electric machines. In contrast to the existing parameter measurement method that applies a voltage pulse excitation to the test motor for flux linkage measurement, the current control method developed in this thesis uses a current pulse with closed loop control. While the voltage pulse method requires a higher sampling frequency for machines with lower time constants, the number of samples available during the transient for a fixed sampling rate can be modified by using a current pulse to measure the flux linkages. In addition, the use of a current pulse in closed loop makes it possible to measure the machine flux linkages at operating points unattainable with the voltage pulse method due to inverter dead time and device drops. This PhD work also develops a current controller design methodology for the developed parameter measurement method, which aids in the automation of the measurement process in a real time system.
This PhD research also develops an experimental method to obtain the static torque angle curves and torque-ripple of synchronous reluctance machines (SynRM). The developed technique is used to study the performance of a SynRM using cold rolled grain oriented (CRGO) laminations against a regular SynRM using cold rolled non-grain oriented laminations. The SynRM using CRGO laminations are designed to have a higher saliency, and thus a higher torque per ampere and higher power factor. This thesis presents a comparative study of the torque performance of the CRGO and CRNGO SynRMs.
This PhD work also develops the mathematical model of an interior permanent magnet synchronous machine (IPMSM) with aligned magnet and reluctance torques. It is a new class of IPMSM also called shifted IPMSM, which is designed to have higher torque for lower magnet volume. The torque characteristics of the shifted IPMSM is different from conventional IPMSMs. For the analysis and operation of this new class of machine, a suitable mathematical model is lacking in the literature. The PhD work develops the mathematical model of the novel shifted IPMSM, and validates it using the experimentally obtained inductance, torque angle and torque-speed curves.
This thesis also develops a mathematical model of a novel hybrid variable flux machine (VFM) having rare-earth magnets in series with AlNiCo magnets for power hardware in the loop emulation applications. In order to emulate the VFM, emulation of the change in magnetization state is crucial. This thesis models the magnetization and demagnetization characteristics of the VFM as look-up tables. The current control method to measure the machine parameter proposed in this thesis is used to measure the flux linkage characteristics of the VFM, and a complete VFM model suitable for power hardware in the loop emulation is developed. The developed model is validated experimentally through the comparison of transient and steady state machine behavior.
The works presented in chapters 2 and 3 are primarily useful in generating the flux-linkages, inductances and torque look-up tables for behavioral model special electric machine models. Such a model is used in advanced model based controls that would not be possible with basic two axis machine models. Moreover, the developed techniques are useful in validation of finite element analysis simulation of special electric machines. Within the thesis, Chapters 4 and 5 used the techniques developed in chapters 2 and 3 to develop and validate the machine models of shifted interior permanent magnet synchronous machine and hybrid variable flux machine. The developed models are useful in drive operations of these motors using advanced control algorithms.