提出了一种采用定子电流和扰动观测补偿的改进预测电流控制(PCC)算法。理论上永磁同步电机PCC具有优异的控制性能,但现实系统中离散采样延时与电磁参数时变等问题使得原理上基于模型的预测电流控制器的控制品质严重恶化。设计了龙伯格(Luenberger)状态观测器实现对定子电流与参数扰动的观测,并应用于补偿和改进经典的无差拍预测电流控制器。将观测到的当前时刻的定子电流替代当前时刻的采样电流用于反馈控制,以补偿采样延时；运行过程中电磁参数的变化使得模型参数失配,其影响以电压扰动的形式被观测出来,并补偿到预测控制输出的电压指令中。仿真和试验结果验证了所提方法的有效性。

An improved predictive current control (PCC) algorithm with stator current and disturbance observation compensation was proposed. Theoretically, the PCC for the PMSM possesses excellent control performance. However, problems including the discrete sampling delay and the timevarying electromagnetic parameters in the practical system cause serious deterioration in the control quality of the modelbased predictive current controller. A Luenberger state observer was designed to observe stator current and parameter disturbances, which was applied to compensate and improve the conventional deadbeat predictive current controller. The sampling current at the present moment was substituted by the observed stator current at the present moment for the feedback control to compensate the sampling delay. The variation of the electromagnetic parameters during operation caused a mismatch in the model parameters, the effects of which were observed in the form of voltage disturbances and compensated to the output voltage reference of the predictive control. The simulation and experimental results verified the validity of the proposed algorithm.

TM 301.2

台达电力电子科教发展计划(DREG2017018)