Abstract: In order to make full use of Permanent Magnet Synchronous Motor（PMSM） reluctance torque, it proposes a Maximum Torque Per Ampere（MTPA） method of PMSM. The discrete predictive model of electromagnetic torque is established first, then it focuses on the internal mechanism and typical problems of MTPA optimization. And then it builds a predictive control cost function, which contains of electromagnetic torque tracking, MTPA optimization as well as the system current limitation. The cost function can adaptively identify the dynamic and steady state condition of PMSM speed control system online. And according to the corresponding identification results highlight the optimization goal, it realizes the global optimal control of PMSM speed control system. Matlab/Simulink results show that the method can improve the efficiency of PMSM speed control system, at the same time it retains some advantage of predictive control, such as multi-objective optimization and high dynamic response. In addition, the method shows a strong robustness when the motor parameters are changed in realizing the MTPA optimization control.

Key words: permanent magnet synchronous motor, predictive control, efficiency optimization, cost function

摘要： 为了充分利用永磁同步电机（PMSM）的磁阻转矩，提出了一种PMSM最大转矩电流比（MTPA）预测控制优化方法。该方法在建立PMSM调速系统电磁转矩离散预测模型的基础上，重点研究了其MTPA优化的内在机理与典型问题。进而构建了一个以电磁转矩跟踪、MTPA优化以及系统限流保护为目标的预测控制价值函数，该价值函数可以在线自适应地识别出PMSM调速系统的动、稳态情况，并根据相应的识别结果重点突出各个优化目标项，从而实现PMSM调速系统的全局最优控制。Matlab/Simulink仿真结果表明，该方法有效地提升了PMSM调速系统的运行效率，同时其保留了预测控制多目标优化、高动态响应等控制优势。此外，在实现MTPA优化控制时该方法对电机参数的变化展现出较强的鲁棒性。

关键词: 永磁同步电机, 预测控制, 效率优化, 价值函数