Model-Free Predictive Current Controller With Ultra-Local Model Utilizing a Novel Observer for Dual Three Phase PMSM

Existing current control approaches of permanent magnet synchronous motor (PMSM) such as deadbeat predictive current control (DPCC) and model predictive current control (MPCC) offer fast dynamic response and certain degrees of robustness. However, they heavily rely on accurate knowledge of machine parameters. While proportional integral derivative (PID) provides certain robustness against parameter mismatch, it requires fine tuning of the controller gain with a trade-off between set-point tracking and disturbance rejection performances. Furthermore, the parameters vary with different operating and environmental conditions, leading to performance degradation. Moreover, external disturbances are difficult to model precisely. To address this issue, this paper proposes a new model-free predictive current controller integrated with a nonlinear disturbance observer (MFPCC-NDOB) for a dual three-phase PMSM. This approach reduces the dependency on parameter accuracy and precise system modeling, thereby enhancing the robustness against parameter mismatch and external disturbances. The proposed MFPCC-NDOB method is evaluated on a 9 kW DTPPMSM using MATLAB/Simulink and compared with DPCC and conventional MFPCC. The simulation results validate that the proposed controller achieves improved tracking accuracy and superior overall performance in practical applications.