永磁同步电机模型预测转矩控制模糊排序法研究

Model predictive torque control for permanent magnet synchronous motor based on fuzzy ranking approach

中文关键词  永磁同步电机  模型预测转矩控制  模糊控制  排序法  权重系数

英文关键词  permanent magnet synchronous motor  model predictive torque control  fuzzy control  ranking approach  weighting factor Citation: LI Yaoh

中文摘要

      针对永磁同步电机(PMSM)模型预测转矩控制(MPTC)中成本函数权重系数难以设计和调节的问题, 以降低逆变器开关频率的多目标控制问题为例, 本文研究了基于排序法的模型预测转矩控制策略. 本文通过优先级的设计解决了排序过程中最优电压矢量解不唯一的问题. 考虑到控制目标重要性并不完全等同, 提出了一种带有缩放因子的排序优化方法, 利用缩放因子来调节控制目标的重要程度. 不同于连续变化的权重系数, 缩放因子的作用效果具有离散分段特性, 因此其调整过程可得到有效简化. 本文进一步提出了基于模糊排序法的模型预测转矩控制策略, 实现了缩放因子的动态优化, 从而更好地适应电机不断变化的运行状态. 仿真结果表明, 与固定权重系数的传统模型预测转矩控制相比, 本文所提控制策略可降低系统的平均开关频率、转矩与磁链脉动, 并可有效抑制动态条件下的转矩和磁链脉动. 实时性实验结果表明, 排序法不会严重降低模型预测转矩控制的实时性.

英文摘要

      Aiming at the problem that the weighting factors of cost function are difficult to design and tuning in model predictive torque control (MPTC) for permanent magnet synchronous motor (PMSM), taking the multi-objective control problem of reducing the switching frequency as an example, the MPTC based on ranking approach is studied. The priority is designed to avoid the problem that the optimal solution of voltage vector is not unique in the ranking process. Considering that the importance of control objectives is not exactly equal, a ranking-based MPTC with scaling factor is proposed, which uses scaling factor to tune the importance of the control objective. Different from the continuous weighting factor, the effect of scaling factor has piecewise characteristic, so its adjustment process can be effectively simplified. Furthermore, a MPTC based on fuzzy ranking approach is proposed to realize the dynamic optimization of scaling factor, so as to better adapt to the changing operating points of motor. The simulation results show that compared with conventional MPTC using fixed weighting factor, the proposed control strategy can reduce the average switching frequency, torque and flux ripple of the system, and can effectively suppress the torque and flux ripple under dynamic conditions. The test results of algorithm execution time show that the ranking approach does not seriously degrade the real-time performance of MPTC.