基于滑模磁链观测器的感应电机模型预测控制

感应电机模型预测控制通过系统状态预测的方法,利用设计好的目标函数直接控制磁链和转矩。该方案易于实现多变量控制和处理非线性约束,但因为预测过程需要使用电机的真实磁链,所以对磁链观测的精度有很高的要求,以滑模磁链观测器替代传统的电流模型和电压模型磁链观测器可大大提高该方案的控制性能。常规的滑模磁链观测器依据电流观测误差设计滑模控制函数,在实现对转子电阻变化的高鲁棒性的同时也引入了定子电阻变化的影响。文章中提出一种基于磁链观测误差的滑模控制函数设计方法,对滑模观测器的性能进行改进。因为磁链的真实值无法直接获取,磁链误差值以近似计算的方式得到。通过仿真和物理实验验证了该方案对电机参数变化的高鲁棒性。

A Sliding Mode Flux Observer for Predictive Torque Controlled Induction Motor Drive

The model predictive control scheme of induction motor predicts the system state and use a designed cost function to control the flux and torque directly. The scheme is easy to add different targets and nonlinear constraints,but has a high requirement on the observation precision of the motor flux. A sliding mode flux observer for model predictive control was proposed in this paper. The conventional sliding mode terms were designed based on current observation error,it realized the high robustness to rotor resistance variation but also introduced in the effect of stator resistance. This paper proposed a method to improve the performance of the sliding mode flux observer by designing sliding mode terms based on flux observation error. As the real value of stator flux can't be acquired directly,the flux observation error was obtained by approximate calculation. Simulation and physical experiments proved the sliding mode observer improve the parameter robustness of predictive torque controlled induction motor drive.