无差拍优化五相永磁同步电机有限集模型预测转矩控制算法

Finite-control-set Model Predictive Torque Control Algorithm With Deadbeat Optimization for Five-phase Permanent-magnet Synchronous Machines Drives

多相永磁同步电机驱动系统在低压大功率、高可靠性的应用场合日受青睐和瞩目。首先研究了适用于五相永磁同步电机系统的传统有限集模型预测转矩控制(finite-control-set model predictive torque control,FCS-MPTC)算法,为了消除d3-q3谐波子空间的低次电流谐波,构造了包含谐波项的目标函数。然后,为了减小FCS-MPTC算法实现时带来的巨大计算量,提出了一种基于转矩与磁链无差拍估算的每个控制周期内电压矢量控制集优化方法。所提算法不仅保持了传统FCS-MPTC算法优越的稳态性能、快速的动态响应和低次电流谐波抑制能力,同时显著减少了数字实现的运算量。最后,将所提算法分别与两种传统FCS-MPTC算法进行半实物实验对比分析,验证了所提算法的正确性和有效性。

The favor and attention in multiphase permanent-magnet synchronous machines （PMSM） drives has substantially increased day by day, due to the potential advantages in high-power with low-voltage and high reliability applications. This paper discussed two traditional finite-control-set model predictive torque control （FCS-MPTC） schemes firstly. In these schemes, the third order harmonic items were included in cost function in order to eliminate the low-order current harmonics caused in d3-q3 subspace. And then, a deadbeat predictive FCS-MPTC scheme for five-phase permanent-magnet synchronous motor （PMSM） drives was proposed to optimize the selection process of the best switching vector in every control interval. The proposed algorithm can bring the following benefits： fast control dynamic response, superior steady-state performance and low-order current harmonic reduction. Furthermore, the proposed scheme can relieve the heavy calculation burden in the traditional FCS-MPTC methods, dramatically. The experimental hard-in-loop （HIL） results have verified the correctness and effectiveness of the proposed scheme.