Non-linear state estimation of PMSM using derivative-free and square-root Cubature Kalman Filter

This paper presents a non-linear state estimation of Permanent Magnet Synchronous Motor (PMSM) with currently developed non-linear estimator. The estimation tool used is Cubature Kalman Filter (CKF). The CKF is a derivative-free online sequential state estimator; compared to other nonlinear filters, it depends on integration for its operation. It inherits the property of original Kalman filter, including derivative-free and square-root filtering for improved stability and reliability. It can minimize errors due to linearization. Despite of having several advantages, CKF is not yet explored in the field of electric drives. Here PMSM is modeled in dq reference frame and for modeling of CKF and its algorithm uses modeling of PMSM in stationary αβ reference frame. The speed and rotor position is estimated and then it is feedback to PMSM for speed control. Simulation results shows the efficiency proposed estimator.