Robust and optimal control of AC machines

Motor control applications such as semiconductor manufacturing machines, robotic arms, electric vehicles, hard-disk drives and home appliances, can be found everywhere. In the past, direct current (DC) brush motors are used extensively for various applications due to the ease of speed and position control. However, the mechanical commutation brushes of DC motors require periodical maintenance, cannot operate in high speed regions and cannot sustain high voltage and high current. With the help of advanced control algorithms, efficient power electronic converters and powerful digital signal processors, all DC motors can now be replaced by alternating current (AC) machines with the same system output performance, but without the problems caused by the commutation brushes. Standard nonlinear control algorithms such as field-oriented control (FOC) can transform the highly nonlinear AC machines into linear systems; nevertheless, the system output performance degrades when the system plant is perturbed and/or external disturbances are present.; In this thesis, the robustness issue of AC machine control is addressed. Three industrial problems of AC machine control are studied. These are, namely, velocity ripple elimination of AC permanent magnet (PM) motor systems, robustness enhancement of indirect field-oriented induction machine drives, as well as the development and control of a linear switched reluctance motor (LSRM) for high precision applications.; Robust control techniques including $H2$ and $H\infty$ optimizations, internal model principle, and two degree of freedom (2DOF) control structure, are employed in the above three problems to design the motor controllers on top of the nonlinear control algorithms, so that the reference transient tracking response, the external disturbance rejection performance and the system robustness could be improved under plant parameter variations and in the presence of external disturbances. Three test beds are developed to perform practical experiments and verify the effectiveness of the proposed motor controllers. The simulation and experimental results reveal that the overall system output performance can be improved using our proposed robust controllers.