Modified Robust Predictive Current Control Design of Windings-Discontinuous-Segmented PMLSM Based on an Improved Analysis Method

A permanent magnet linear synchronous motor (PMLSM) with discontinuous-segmented windings, driven by a distributed control system, is subject to high current sampling noise and strong back-back electromotive force disturbances. Therefore, an improved robust predictive current control (RPCC) algorithm is proposed to simultaneously achieve better disturbance rejection and noise suppression performance while maintaining a suitable stability margin. First, an improved analysis method is proposed based on frequency response of a fully equivalent unity-negative feedback multi-input-multi-output (MIMO) system for comprehensive performance and stability metrics of the current closed loop. Contrary to general knowledge, for robust deadbeat predictive current control (DPCC) that combines linear extended state observer (LESO), the equivalent integrator in LESO does not reduce the deadbeat tracking characteristics of DPCC for current commands. However, as the bandwidth of LESO decreases, the excess stability margin limits the disturbance rejection and noise suppression performance of the current closed-loop system. Therefore, an improved RPCC algorithm is proposed to overcome this challenge by adding a damping term and a gain adjustment factor to the feedback controller. The correctness of the analysis method and the effectiveness of the improved control method are verified through sweep frequency experiments and time domain response experiments in a PMLSM prototype.