Speed-Sensorless Control System of Bearingless Induction Motor Based on the Novel Extended Kalman Filter

To improve the performance of bearingless induction motor drives with speed sensorless vector control system, a novel extended Kalman filter with a series structure is proposed. The on-line calculation of motor parameters is achieved by extending the easily changing motor parameters to the system model to be the state vector to be identified. The accurate identification of the motor speed can be realized by feeding back the obtained parameter value into the algorithm, which is conductive to reduce the impact of motor parameter variation on speed estimation accuracy. The order of system model matrix can be decreased by employing the series structure extended Kalman filters, and also the computational load and complexity of digital chips in practical applications would be reduced. Thecomparision of estimated speed error between the traditional extended Kalman filter and the novel extended Kalman filter in the case of motor parameters are changed is completed by simulation and experiment. The results demonstrate that the novel Kalman filter can effectively reduce the impact of parameter variation on the estimation accuracy. © 2018, Electrical Technology Press Co. Ltd. All right reserved.