Improving Extended Kalman Filter for Speed Sensorless Vector Control of Induction Motor

LI Xu-chun,ZHANG Peng,ZHAO Fei
DOI: https://doi.org/10.3969/j.issn.1007-449x.2013.09.004
2013-01-01
Abstract:Classical extended Kalman filter(EKF) poses challenges like high dimensional iterative matrixes and a large amount of computation.To tackle these problems,a speed sensorless vector control method of induction motor was proposed,using a new EKF-based observer.The quadrature-axis rotor flux is equal to zero with rotor field oriented control.With this fact,this method simplified mathematical model of induction motor by controlling the flux amplitude constant below the rated frequency,and by using the calculation method of slip angular velocity in indirect vector control for reference.A three-order EKF-based observer was constructed to estimate the flux and the rotor speed simultaneously,by selecting the direct-axis and quadrature-axis components of stator current and the angular velocity of stator field as state variables.Compared with the classical full-order EKF-based observer,the improved observer reduced the orders of model and the complexity of system,and lightened the computational load.Simulation and experimental results show the validity and the practicability of the constructed speed sensorless vector control system of induction motor,in which the flux and the rotor speed are accurately estimated by the improved observer.
What problem does this paper attempt to address?