In the operation of brushless DC motors, a method to observe angular velocity from hall sensor signals is widely used. However, the observation accuracy deteriorates due to misalignment of hall sensors and magnetic poles on the rotor. To solve this problem, Kalman filtering based on the observed signals is effective to estimate angular velocity. In order to improve the estimation performance, the proposed method adopts a parameter setting for the variance of the measurement noise in Kalman filter algorithm, which is set to be depending on the estimated angular velocity. This paper shows that the variance of the measurement noise is obtained through a simple preliminary experiment and the effectiveness of the proposed method is verified by experimental results.
