Observer-based sensorless control of synchronous reluctance motor

Synchronous motors are electrical machines that can generate mechanical energy with the reluctance moment principle. Since it does not contain any winding and magnet in its rotor, it has a cheap and robust structure. since it does not contain any winding and magnet in its rotor. For these reasons, although its discovery dates back to earlier times, it has recently become popular due to advances in power electronics and control methods. Synchronous reluctance motor cannot be directly connected to the grid for constant speed applications such as asynchronous motor, and the motor requires a driver circuit. The motor driver circuit needs the angular position of the rotor to make the necessary switching. At this point, control methods are divided into position sensor and sensorless. In this thesis, synchronous reluctance motor was controlled without sensor. In the sensorless control method, the rotor angular position is estimated with the principle of active flux concept previously proposed for all alternating current machines. Unlike previously proposed methods of estimating rotor position with active flux vector, the state-space equation set is created by splitting the active flux vector into differential equations. Active flux vector components are estimated by designing the state-space equations Unscented Kalman Filter (UKF). The angular position of the rotor has been estimated thanks to the components. Unlike other active flux and rotor angular position estimation methods, it does not include any integrators or Phase Locked Loops (PLL). The axis transformation is made with the rotor angular position value estimated with the UKF and the transition is made to the Park axis. Synchronous reluctance motor dynamic equations created in the park axis Extended Kalman Filter (EKF) was designed. With the designed EKF, rotor angular speed and load torque can be predicted. The estimated rotor angular velocity was compared with the estimation values in other estimation methods, and it was observed that the rotor angular velocity error estimated by EKF was 70% less than the other methods. UKF and EKF algorithm work together. Because the rotor angular velocity estimated in EKF is accepted as constant within UKF and transactions are made. The synchronous reluctance motor was controlled by applying field oriented control methods for the estimated rotor angular position, speed and load torque values. The mentioned control steps were checked in the test setup after the simulation study in computer environment. Silicon carbide (SiC) type mosfets are used in the driver board in the test setup, which performs better than other power switches. In this way, the performance of the circuit and system has been increased.