Abstract

In this thesis, online flux linkage estimators for synchronous machines (SMs) are proposed. The proposed flux estimators are generally applicable to all types of synchronous machines and are designed to consider the nonlinearities caused by cross-coupling and magnetic saturation in the machines. Additionally, the estimators are designed without using filters designed in the frequency domain, thus avoiding phase and magnitude distortions in transient state. Existing estimators primarily determine their estimation performance in both transient and steady states based on the accuracy of the nominal parameters. However, the proposed estimators assume the nonlinear flux term as a ramp disturbance signal or estimate parameters online and update them to the actual values, thereby reducing the impact of parameter inaccuracies. This improves the flux estimation performance under various operating conditions. The validity of the proposed estimators is demonstrated through comparisons with existing estimators with finite element method (FEM)-based simulations of a 35 kW Interior Permanent Magnet Synchronous Motor (IPMSM) provided in MATLAB/Simulink.