Contribuições ao acionamento e controle sensorless aplicado ao motor de indução bifásico assimétrico

Abstract

This doctoral thesis presents the development of sensorless methods aiming the rotor speed control of asymmetrical two-phase induction motors. Initially, the machine mathematical model is presented. From the machine model, two methods for identification of electrical parameters of the induction machine are presented. The first method is based on the classical tests for electrical parameter estimation on three-phase induction motors, while the second method uses a RLS algorithm for the electrical parameters identification. Moreover, a vector control scheme applied to asymmetrical induction machines is presented. The influence that the machine asymmetry impacts on the field indirect vector control rotating at synchronous speed is analyzed. Simulation results show the coupling and the oscillations on the stator currents. The rotor speed estimation is carried out firstly by a MRAS estimator applied to single-phase induction motor drive. Furthermore, three rotor speed estimation methods based on continuous and discrete time sliding mode observers are proposed. The proofs of stability and convergence of these algorithms are developed and presented. The limits for the switching gains of the sliding mode observer are presented for continuous and discrete time. These limits are distinct and are highlighted in this study. Besides, a geometric modulation technique for three-leg voltage source inverters driving two-phase machines is presented. The rotor speed control is carried out from two schemes. The first method uses a discrete-time PI controller, while the second scheme uses a RMRAC controller. Simulation and experimental results are presented to validate the proposed techniques.