RLS-based deadbeat predictive current control for dual three-phase segmented powered linear motors

Deadbeat predictive current control (DPCC) is an effective model-based motor control method. However, due to the unbalanced inductance and parameter variations of the segmented powered linear motor stator, the conventional model of linear motors is not accurate, which ultimately affect the performance of the control. This paper proposes a novel DPCC based on the recursive least squares (RLS) method to identify the parameters of the dual three-phase segmented powered linear motor (SP-LM) model. First, the influence of unbalanced inductance caused by the segmented motor stator and parameter variations of the conventional DPCC are analyzed. Second, a discrete RLS model of the dual three-phase SP-LM is established, which is a common model for both linear induction motors (LIMs) and linear synchronous motors (LSMs). Finally, the model parameters are identified by the RLS method and the deadbeat principle is used to predict the current. The proposed method effectively eliminates the influence of unbalanced inductance and the parameter variation, improves the current control performance and reduces the thrust fluctuation. Experiments based on hardware-in-the-loop verify the proposed method.