Sensorless control of permanent magnet synchronous machine

Permanent magnet synchronous motors (PMSM) are commonly used in the aerospace industry, for example as actuators for flying surface control of aircraft, thus replacing hydraulic technology, or as the driving unit for attitude control of satellites. The primary advantages of the (PMSM) drive are high dynamics and high energy efficiency. Controlled electric drives without mechanical speed and position sensors at the motor shaft are attractive for their low cost and high reliability. The sensorless control of the PMSM uses the estimated speed and the estimated rotation angle to replace the measured speed and rotation angle. The speed and position estimators can be divided into two groups. The first group is high-frequency signal injection-based methods, which are effective at low speeds. The second group is back electromotive force (back-EMF) based estimator methods, which are effective at high-speed operation. This project aims to study and propose a sensorless control method based on a back-EMF estimator in simulation with Matlab/Simulink before validating it experimentally using a dSPACE1104 board. The provided control solution will be implemented on a three-phase synchronous permanent magnet motor to avoid the use of position sensors (rotary or linear) necessary to correctly excite the phases so as not to disperse the electrical energy fed into heat and to overcome possible wear or maintenance required by the sensors. This report initially presents the properties of the PMSM motor studied. In the second chapter is shown the implementation of the modelling of the motor. A PI (Proportional- Integral) controller is then applied for current and speed control (Chapters 3 and 4) and the results of the system's simulation are shown. After a bibliographic study of some observers-based control methods, in Chapter 5 the LADRC controller and the ESO observer are proposed and simulated. The aim is to replace the PI controller with first-order LADRC controller, which has better performances in disturbance rejection. A comparison between the behaviour of the PI and LADRC controllers is presented at the end of the section. Finally, in Chapter 6 the implementation of sensorless control through an ELO (Extended Luenberger Observer) and the PLL (Phase-Locked-Loop) rotor position tracker is presented and tested using the dSPACE1104 board.