摘  要：为了提高传统直接转矩控制系统的性能，分析了模糊控制和交互式模型参考自适应辨识器的基本工作原理。采用模糊控制器取代原来的磁链、转矩控制器，对不同的磁链误差、转矩误差选择不同的电压矢量，可以减小转矩误差；同时采用了一种基于交互式模型参考自适应理论的无速度传感器，可以通过变换参考模型和可调模型来辨识速度和定子电阻。结果表明，基于交互式模型参考自适应的模糊直接转矩控制系统的磁链和转矩的脉动小，提高了低速时的性能，同时可以实现对转速和电阻的辨识。因此，基于交互式模型参考自适应的模糊直接转矩控制与传统的直接转矩控制相比具有很大的优越性。
关键词：直接转矩控制（DTC）；模糊控制；模型参考自适应（MRAS）；交互式模型参考自适应

Abstract: The principles of fuzzy control and mutual model reference adaptive identifier were analyzed to improve the performance of traditional direct torque control (DTC) system, a fuzzy controller was adopted to replace the original flux controller and torque controller, it can select different voltages according to different flux errors and torque errors, so the torque error was reduced. A speed-sensor based on the theory of mutual model reference adaptive system (MRAS) was also adopted, the speed and stator resistance can be identified by varying the reference model and adjustable model. The result of analysis shows that the fuzzy DTC system based on mutual MRAS can solve the problems of torque and flux ripple, improve performance of the poor low-speed. Therefore, the fuzzy DTC system based on mutual MRAS is prior to the traditional DTC system.
Key words: direct torque control (DTC); fuzzy control; Model reference adaptive system (MRAS); mutual model reference adaptive system