Model-free deadbeat predictive current control of NIPMVMs considering dead-time effect

This article proposes an adaptive model-free deadbeat predictive current control (MF-DPCC) algorithm for achieving high-performance current control of non-contact integrated permanent magnet vernier motors (NIPMVMs). To address the issue that traditional extended state observer-based deadbeat predictive control (ESO-DPCC) still relies on motor models and parameters, an ultralocal model is introduced to design a model-free current controller. Furthermore, to address the issue that initial gain of the controller depends on the inductance parameter, a gain-adaption algorithm based on the deadbeat characteristic is proposed. In addition, considering the effect of dead-time, a dual-amplitude signal with extremely low duty and amplitude levels has been adopted to improve the adaptive precision of the gain and to achieve deadbeat control of NIPMVMs. The validity of the proposed MF-DPCC is confirmed through experiments conducted on an NIPMVM platform.