چكيده به لاتين
Nowadays, due to many benefits of flywheel energy storage system such as nonenvironmental pollution and the frequency of unlimited charging and discharge, this system has many applications in different industries. The use of synchronous reluctance motor, because of its high efficiency and more comfortable control principles, it can be a suitable option for use in the flywheel energy storage. Since the past years, the use of linear control method has been widespread, but today, due to the significant progress of microcontroller computational power, the use of new control methods have been proposed. Model-based predictive control is one of the most practical and attractive approaches that today has been widely used in motor drives and power electronic converters. In general, this control method is divided into two ways that consist of continuous and finite-state. In the finite set predictive control, all computations are done online, which increases the volume of calculations, but in the constant set, most of the computations are done offline, thus the computing burden will be more decrease. In the design of continuous control set, sampling time, and prediction horizon have a significant effect on the dynamic performance of the system. This thesis is suggested that according to the dyinamic response of the system the sampling are selected and according to the chosen sampling time, the appropriate predection horizon is selected which can be guaranteed the system stability. For Linear Time Varation (LTV) system, a new structure is proposed that it is not required to apply nonlinear algorithms or online calculations, and only with the use of model error, improves the controllers performance in LTV and nonlinear system moreover if unexitency parameter mismatch and nonlineariy in system model, the proposed structure have not any effect in control system performance.
