A predictive current controller is established based on the system model and therefore suffers from the problems such as the parameter mismatches, the digital delay and the external disturbances. In order to tackle these problems, an observer-based robust predictive current control strategy is proposed in this paper, which employs an extended state observer to estimate the disturbances, currents and applies the augmented system model for the design of the controller. It is generally formulated and applicable to any AC motor drive system, as long as the prerequisites are satisfied. Furthermore, the proposed control scheme is proven to be input-to-state stable with a proper design of the controller and the observer. The effectiveness of the proposed control strategy is moreover verified with numerous experiments on a dSPACE system under various test scenarios. The comparative studies between the proposed method and the predictive current control approaches, i.e. the conventional deadbeat control and the continuous control set model predictive control, are conducted at different operating conditions.     «

A predictive current controller is established based on the system model and therefore suffers from the problems such as the parameter mismatches, the digital delay and the external disturbances. In order to tackle these problems, an observer-based robust predictive current control strategy is proposed in this paper, which employs an extended state observer to estimate the disturbances, currents and applies the augmented system model for the design of the controller. It is generally formulated a...     »