Abstract
The major purpose of this dissertation lies in the development of some advanced circuits and control techniques to promote the ability of the switched reluctance motor (SRM) in high-performance applications. After surveying the existing converter circuits, an experimental SRM drive is established. In developing a model for the SRM, its nonlinear winding inductance is estimated by curve fitting according to measurements. Following that, a simulation environment for the overall SRM drive is built up using POWERSYS and SIMULINK. In order to improve the performance of the SRM at high speeds, an active voltage boosting circuit is developed to reduce the effects of the back electromotive force (EMF) and the winding inductance on the current response. The circuit operation and the switching control of the proposed voltage boosting circuit are described in detail. The simulation and experimental results show that the winding current tracking response and hence the torque generating capability are significantly enhanced through applying the developed voltage boosting circuit. Since the winding current tracking trajectory and the inductance profile of an SRM are far from the predicted shapes, its torque generating characteristics are difficult to specify and optimize quantitatively. To improve this, according to the observations about the effect of the commutation instant on the torque generating characteristics, an intelligent tuning approach is proposed. The minimization of the motor-drawn current is employed as a performance index to equivalently yield the maximum torque per ampere (TPA). Through utilizing the proposed tuning mechanism, the motor conversion efficiency and the speed dynamic response can be considerably improved. As to the speed control of the SRM drive, the quantitative and robust two-degree-of-freedom (2DOF) control approaches have not been addressed. In this dissertation, the estimation of a dynamic model and the quantitative design of a robust 2DOF speed controller using the inverse model concept are presented. According to the estimated dynamic model and the prescribed load regulation speed control specifications, the feedback controller is first designed. Then, a command feedforward controller is designed to let the speed tracking response follow the response defined by a reference model. As variations in system parameters or operating conditions occur, a robust controller (RC) is added to preserve the model following speed tracking response as far as possible, and at the same time the speed variation due to the load torque change prescribed in the design stage can be reduced further. In designing the proposed RC, the effect of system transport delay on the closed-loop stability is also taken into account. Finally, a 2DOF controller using a fuzzy adapted inverse model is developed. As generally recognized, to attain a perfect speed tracking response, the command feeedforward controller should be exactly equal to the inverse of the actual motor drive model all the time. To achieve this goal, the inverse model employed in the command feedforward controller is adaptively tuned using a fuzzy control technique. Since the command feedforward control scheme is irrelevant to the speed load regulation response, a disturbance cancellation robust controller considering system transport delay is added to reduce speed variations due to load torque disturbances, and meanwhile the speed tracking response can be improved still further.