Robust Current Predictive Control for Dual-Stator Permanent Magnet Synchronous Motors based on Extended State Observer design

Abstract

With the quickly development of the new energy industry, Relative to standard permanent magnet synchronous motors (PMSMs), dual-stator permanent magnet synchronous motors (DSPMSMs) boast higher output torque and a wider stable speed operation range, which has earned them extensive academic attention. However, classic deadbeat predictive control (DPC) schemes algorithms for DSPMSM suffer from poor robustness against parameter mismatches, which significantly degrade control performance when motor parameters such as inductance and resistance vary during operation. In response to this issue, a deadbeat predictive current control (DPCC) strategy without manual parameter tuning, which is based on an extended state observer (ESO), is put forward in this paper. To begin with, this work establishes the mathematical model for the DSPMSM, and completes the design of traditional DPCC as well as deadbeat predictive speed control (DPSC) algorithms. After that, the influence of parameter deviations on control performance is analyzed, and the conclusion is drawn that inductance variations account for most of the current tracking errors. On this basis, second-order ESO structures are developed for both the current loop and speed loop to observe and compensate for lumped disturbances including internal parameter perturbations and external load disturbances in real time. Finally, simulation verification is conducted in Simulink under different parameter mismatch conditions. This result indicates that the proposed ESO-based DPCC reduces current and speed fluctuations.