Robustness Evaluation of a Sensorless Control of Synchronous Reluctance Motor with a Kalman Filter & a Luenberger Observer

This paper investigates the simulation of a sensorless speed control method with estimates the position and the speed of a synchronous reluctance machine (SyncRM). Speed control implements Field Oriented Control (FOC)., fed with estimated speed and position. The method combines two different estimators. First., a Luenberger observer estimates the load torque of the machine from the mechanical model of the machine., knowing the value of the inertia coefficient and the dynamic friction coefficient. Then, a Kalman filter estimates the speed and the position of the machine., from differential state equations resulting from the analytical (electrical) model of the machine in the rotational reference frame, taking benefit of the estimated load torque. The robustness of the method was verified by Matlab/Simulink simulation for different speed and torque profiles., against parameter errors of the electrical (inductance and resistance) and mechanical (inertia and dynamic friction) models of a 5kW machine.