一种改进的永磁同步牵引电机低开关频率模型预测直接转矩控制策略

An Improved Low Switching Frequency Model Predictive Direct Torque Control Strategy for Traction Permanent Magnet Synchronous Motor

为了进一步提高永磁同步牵引系统的效率及其动静态性能,文中提出一种改进的模型预测直接转矩控制策略。在有限集模型预测控制中引入直接转矩控制的滞环概念,将逆变器输出电压矢量的寻优标准从传统的跟踪误差最小,优化为在滞环内延展步长最长,以降低逆变器的开关频率,提高系统效率。同时,通过等分开关周期的方法构造32个虚拟候选矢量,提高控制器的跟踪精度和动态性能。最后,在低成本DSP硬件平台上实现该控制策略,实验证明,相对于传统的单矢量模型预测直接转矩控制,所提算法在降低逆变器开关频率的同时,能有效提升电机的动静态性能。

In order to improve the efficiency and control performance of traction PMSM,this paper proposed an improved model predictive direct torque control strategy.The concept of hysteresis band of traditional direct torque control was introduced into the finite set model predictive control.The optimization standard for selecting output voltage vector was optimized from minimizing the tracking error to maximizing the extension step in hysteresis band,so as to reduce the switching frequency and improve system efficiency.At the same time,32 virtual candidate vectors were constructed by equally divide the switch period,which improves the tracking accuracy and dynamic performance of the controller.Finally,the control strategy was implemented on the low-cost DSP hardware platform.The experiments show that compared with the traditional single vector model predictive direct torque control,the proposed algorithm can effectively improve the dynamic and static performance of the motor while reducing the switching frequency of the system.