Model Predictive-Position Sensorless Control of PMSM with Non-Sinusoidal Back-EMF

Model predictive control (MPC) has proven to be an efficient control technique for torque ripple minimization in permanent magnet synchronous motors (PMSMs). The torque ripple in practical PMSM is primarily caused by the non-sinusoidal back-EMF owing to the exact shape and structure of the rotor magnets and the equivalent non-sinusoidal flux distribution within the air gap. The predictive control techniques use the rotor flux lookup table (LUT) to estimate the motor output torque under non-sinusoidal back-EMF. This work proposes a rotor position estimation technique by using the rotor flux LUT and by reverse calculating the α and β components of the rotor flux components from the measured motor phase currents and inverter control signals. The proposed technique estimates the rotor angle irrespective of the initial position without requiring additional sensors. The sensorless control is modeled and tested in the simulation environment for a PMSM of 7 kW rated power, and the design is validated.