Permanent magnet synchronous motor sensorless speed control with rotor position estimation

Traditional control methods of the Permanent Magnet Synchronous Motor (PMSM) use physical sensors to take the position and speed information required in the control. This, not only increases the total cost of the system, but also provides exposure to potential errors and problems that may arise from position and speed sensors such as encoders and resolvers. Possible errors are the tendency to produce Electromagnetic Interference (EMI) that may occur on the sensor cables and errors that may be caused by physical factors such as dirt, moisture and temperature. The use of sensors also greatly increases the size of the machine. In sensorless control, the motor itself acts as a sensor and eliminates the need for a mechanical sensor, resulting in cost reduction. There are two basic methods to determine the rotor position. These are model-based methods and saliency-based methods. In this thesis, the position and speed sensorless control of the PMSM is examined, and a model based, Normalized Least Mean Squares (N-LMS) method is proposed for the sensorless control of the motor. In the method, speed and position estimation was made by using dq axis voltage equations. The simulation of the control of the motor with Field Oriented Control was made by using Matlab / Simulink. Since the applied sensorless control method is an LMS-based method, it can be applied successfully in wide speed ranges. In the study, the success of the method at different speed and load of the motor was tested with simulation studies and high accuracy results were obtained.