Deadbeat predictive current control (DPCC) is well known for its excellent dynamic performance, which requires a sufficient voltage margin to achieve the theoretical tracking delay. However, there is still potential to improve its dynamic performance under conditions of insufficient voltage margin, such as at high speeds and under heavy loads. To address this, this article proposes a modified DPCC based on an exact discretized model of the interior permanent magnet synchronous motor (IPMSM). This model accounts for the zero-order hold effect of the inverter output voltage, providing better accuracy in long-step predictions at the cost of increased computational overhead. Additionally, the motor model is incrementalized to eliminate the need for rotor flux linkage. The proposed method extends the prediction horizon under insufficient voltage margin conditions to cover the entire dynamic process. The stator voltage vector is designed to lie on the hexagonal boundary of space vector pulsewidth modulation (SVPWM) and remain stationary in the $\boldsymbol {\alpha \beta }$ -reference frame throughout the entire dynamic process, thereby shortening the response time through quasitime-optimal control. Furthermore, a simplified iterative method is introduced to reduce the computational cost of voltage calculation, with the iteration number set to three to balance control performance and computational burden. Finally, experimental results validate the rationale for the iteration number selection and demonstrate the effectiveness of the proposed method in improving the dynamic performance of both current and torque compared to the conventional minimum-phase-error voltage limiting strategy. Nevertheless, despite the introduction of the simplified iterative method, the overall computational burden still limits the proposed method to systems with high-performance requirements and sufficient computational resources.
A modified deadbeat predictive current control with improved dynamic performance under insufficient voltage margin for IPMSM
Zhaokai Li
2025-01-01
Abstract
Deadbeat predictive current control (DPCC) is well known for its excellent dynamic performance, which requires a sufficient voltage margin to achieve the theoretical tracking delay. However, there is still potential to improve its dynamic performance under conditions of insufficient voltage margin, such as at high speeds and under heavy loads. To address this, this article proposes a modified DPCC based on an exact discretized model of the interior permanent magnet synchronous motor (IPMSM). This model accounts for the zero-order hold effect of the inverter output voltage, providing better accuracy in long-step predictions at the cost of increased computational overhead. Additionally, the motor model is incrementalized to eliminate the need for rotor flux linkage. The proposed method extends the prediction horizon under insufficient voltage margin conditions to cover the entire dynamic process. The stator voltage vector is designed to lie on the hexagonal boundary of space vector pulsewidth modulation (SVPWM) and remain stationary in the $\boldsymbol {\alpha \beta }$ -reference frame throughout the entire dynamic process, thereby shortening the response time through quasitime-optimal control. Furthermore, a simplified iterative method is introduced to reduce the computational cost of voltage calculation, with the iteration number set to three to balance control performance and computational burden. Finally, experimental results validate the rationale for the iteration number selection and demonstrate the effectiveness of the proposed method in improving the dynamic performance of both current and torque compared to the conventional minimum-phase-error voltage limiting strategy. Nevertheless, despite the introduction of the simplified iterative method, the overall computational burden still limits the proposed method to systems with high-performance requirements and sufficient computational resources.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
