Model Predictive Current Control for PMSM Drives with Low Parameter-dependent Model

Abstract:
In conventional model predictive current control (T-MPCC) method, the motor parameters are subject to perturbation errors due to temperature increase, magnetic saturation, or other nonlinear factors, making the motor parameters inaccurate. Inaccurate motor parameters further lead to inaccurate current prediction models, which adversely affects the motor control effectiveness. In this paper, a low parameter-dependent model control method for PMSM drives is proposed to solve the effects of perturbation errors in motor parameters. The method is suitable for surface-mounted permanent magnet synchronous motor (SPMSM), and the parameter values of the inductance and magnetic flux are extracted from the error between the predicted currents of the q-axis in the ideal condition without parameter perturbation and in the actual operating condition. Then, bring the extracted inductance and magnetic flux parameter into the current prediction model for real-time updating. As verified by simulation, the method can reduce the dependence of the current prediction model parameters, and also can enhance the robustness of the drive system parameters.