Contribuição ao controle sensorless de motores síncronos de ímãs permanentes

Abstract

This thesis presents contributions to sensorless control for permanent magnet synchronous motors. A maximum torque per ampere strategy is used to generate the current control reference in synchronous reference frame. When this type of strategy is used this gives better performance in operations up to rated speed can ve achieved. Two discrete time domain current control proposals are presented for motor applications using microcontrollers. The first technique utilizes a disturbance observer associated sliding mode controller to minimize the effect of matched and unmatched disturbances, as well as variations and uncertainties parameters. The second strategy provides new conditions of linear matrix inequalities for the digital redesign of proportional integral controllers applied to plants described by first order models affected by a sampling period delay in the implementation of digital control. The current control technique presented require the knowledge of the rotor speed/position of the motor. This information that can be obtained through mechanical sensors or rotor speed/position estimation algorithms. Due to the fact that mechanical sensors reduce the robusteness of the control system, three rotor speed estimation techniques were proposed in this work. The first strategy uses the observation of the counter electromotive force of the motor through the use of a disturbance observer, and from this information a speed adaptation law is proposed in the stationary reference frame. The second technique uses a model reference adaptive system. A linearized model is necessary for the development of this technique. In order to increase the robustness of the proposed algorithm the proportional-integral regulator used in classical methods was replaced by a sliding mode regulator. The third estimator, in the discrete time domains, uses the observation of the currents in the discrete time domain to develop a rotor speed adaptation law in the synchronous reference frame. The stability analysis and convergence of the controllers and estimators are presented. Simulation and experimental results validate the proposed algorithms.