Sensorless Synchronous Reluctance Motor Drives: A Projection Vector Approach for Stator Resistance Immunity and Parameter Adaptation

The paper presents a general projection vector framework for the analysis of flux and position observers applied to sensorless control of synchronous reluctance machines, with emphasis to parametric errors sensitivity. The stator resistance immunity property of Adaptive Projection vector for Position error estimation (APP) technique is demonstrated, in maximum torque per ampere (MTPA) conditions. Out of MTPA, additional resistance adaption is devised for accurate estimation of stator flux and torque. Alternatively, inductance adaptation might be preferred to resistance's, when dealing with inaccurate motor flux-maps. Inductance adaptation is shown to decrease the steady-state position error. All proposed APP observers with adaptation techniques are subjected to stability analysis. The merit and the feasibility of the proposed scheme is experimentally demonstrated on a 1.1 kW synchronous reluctance (SyR) machine test-bench.