The position update time increases when the motor with a low-resolution position encoder runs at a low speed. Therefore, a considerable speed calculation delay is generated, and the bandwidth of the speed controller is limited. This paper proposes an adaptive extended Kalman observer, which observes the motor speed and position information through the encoder pulse output and current sampling, and updates the position measurement noise variance in real-time to reduce the motor position and speed estimation error. It is compared with the standard Kalman observer, and the superiority of the proposed adaptive extended Kalman observer is verified through simulation and experiment.

Kalman observer,low-resolution position sensor,low-speed control,speed estimation,permanent-magnet synchronous motor