中文摘要：

针对异步电机矢量控制系统受电机参数变化影响大的问题,采用模型参考自适应系统（MRAS）在线辨识电机转子时间常数。为了增强辨识方法对定子电阻的鲁棒性,提高辨识精度,根据转子磁链模型构造无功功率模型建立MRAS,采用Popov超稳定性理论设计自适应规律,得到基于无功功率模型的转子时间常数辨识方法,抑制定子电阻变化对辨识精度的影响,将该转子时间常数辨识方法应用于异步电机矢量控制系统。仿真结果证明：MRAS有效地提高了矢量控制系统性能,转子磁链轨迹趋于圆形,电流谐波分量明显减少;转子磁链观测误差明显降低,其平均绝对误差相对值为2.87%;转矩脉动大大降低,其平均绝对误差相对值为2.09%。

英文摘要：

According to the fact that performance of vector control（VC）system is greatly affected by the change of motor parameters,model reference adaptive system（MRAS）was applied to identify rotor time constant on-line.According to rotor flux model,reactive power model was built to establish MRAS,and Popov＇s hyperstability theory was applied to design adaptive rule.The method of identifying rotor time constant based on reactive power model was proposed in order to improve robustness against stator resistance and identification accuracy,which reduces the influence of change of stator resistance.The method was applied in VC system of induction motor,and simulation results prove that it improves the performance of VC system effectively.The rotor flux depression and current harmonious decomposition are reduced remarkably,the observation error of rotor flux is reduced sharply to 2.87%and the error of torque is rippled greatly to 2.09%.