Derivative Free Square-root Cubature Kalman Filter for Nonlinear Brushless DC Motors

This paper presents a nonlinear square-root estimation scheme for brushless DC (BLDC) motors. The cubature Kalman filter (CKF) is the main estimation tool for the presented approach. The CKF is a recently proposed estimator for highly nonlinear systems and its efficacy has been verified on several applications. The square-root version of the CKF is preferred over the conventional CKF for real-time applications. Despite of having several advantages over other nonlinear filters, the CKF has not yet been explored for state estimation of electric drives in the electric drives community. In this paper, we present a square-root CKF for the speed and rotor position estimation of a highly nonlinear and high fidelity BLDC motor, these estimated speed and rotor position are then fed back to control the speed of the BLDC motor. The efficacy of the presented approach for low and high reference speeds, and in the presence of parametric uncertainties, is demonstrated by real-time experiments.