Sensorless Control Scheme for PMSM Drive via Generalized Proportional Integral Observers and Kalman Filter

Since conventional linear extended state observer (LESO) can only cope with the zero-order disturbance, the sensorless control scheme based on LESO lacks the ability to accurately estimate the back electromotive force (BEMF), and the estimation accuracy of rotor position and speed is limited. In this article, a new composite sensorless control scheme for permanent magnet synchronous motor drive via generalized proportional integral observers (GPIOs) and Kalman filter (KF) is proposed. First, a GPIO is designed to estimate the BEMF, which can achieve higher estimation precision compared with the conventional LESO. Moreover, the proportional-integral type loop filter is generally employed by quadrature phase locked loop (QPLL). It suffers from weak immunity to time-varying disturbances. To this end, a GPIO-aided QPLL (GQPLL) is designed to synchronously estimate the position, speed, and lumped disturbance of the drive system. Compared with the conventional QPLL, the GQPLL exhibits a better disturbance rejection ability and a higher estimation accuracy. In addition, in order to attenuate the possible measurement noise and reduce the observer noise sensitivity, a composite KF based GPIO (KF-GPIO) for BEMF estimation is proposed. Finally, experimental results are presented to demonstrate the feasibility and effectiveness of the proposed sensorless scheme.