Online-Parameter-Estimation-Based Control Strategy Combining MTPA and Flux-Weakening for Variable Flux Memory Machines

This article proposes a novel online-parameter-estimation-based control strategy for high efficiency and wide speed operation of variable flux memory machine (VFMM), which combines maximum-torque-per-ampere and flux-weakening (FW) technologies. The PM flux linkage is estimated from a newly developed linear active-disturbance-rejection-based feedforward decoupling (LADR-FFD) current controller, while the dq-axis inductances are estimated by employing the recursive least squares algorithm. In the proposed strategy, VFMM is controlled by using only two magnetization states (MSs), namely highest and lowest MSs, to avoid frequent MS manipulation. In the high-speed region of each MS, a continuous d-axis current is applied to stator windings to weaken the air-gap flux for further speed range extension by combining the feedforward and feedback FW approaches. The design procedure of the LADR-FFD current controller and the online estimation methods of the PM flux linkage and dq-axis inductances are first introduced. The proposed online-parameter-estimation-based control strategy is subsequently interpreted with the help of formulas and diagrams. The feasibility and effectiveness of the proposed control strategy are verified through experimental measurements on a hybrid magnetic circuit VFMM prototype.