基于鲁棒模型的航空交流感应电机预测转矩控制

航空领域对驱动电机的动态、稳态及鲁棒性能提出了更为苛刻的要求。近年来，有限集模型预测控制（FCS-MPC）以其动态响应快、稳态性能好（同开关频率下）等诸多优点成为电机驱动领域的国际研究前沿。然而，预测模型依赖于电机参数，鲁棒性差。特别地，在感应电机的FCS-MPC中，模型失配引起电磁转矩预测误差，进而降低系统性能甚至导致系统失稳。针对此问题，提出了一种基于鲁棒模型的感应电机预测转矩控制。在定子电流预测方面，在传统开环模式的基础上，建立了闭环模式的定子电流预测方程，研究了其稳定性及参数设计方法。在定子磁链矢量预测方面，在传统电压模型预测法的基础上，建立了融合电压模型及电流模型的离散混合预测模型，基于比例积分调节器调节电压模型到电流模型的平滑切换。感应电机的电磁转矩则由闭环模式预测的定子电流和定子磁链计算得到。在一台2.2 kW感应电机实验平台上对本文提出的算法进行验证，实验结果证明了该方法的有效性。

Robust model based-predictive torque control of aviation AC induction motor

The dynamic, steady-state and robust performance of the drive motor are more demanding in aviation field. In recent years, Finite Control Set-Model Predictive Control (FCS-MPC) has become an international research frontier in the field of motor drive because of its fast dynamic response and good steady-state performance (at the same switching frequency). However, the prediction model depends on the motor parameters and has poor robustness. In particular, in FCS-MPC of the induction motor, the mismatched model causes the prediction error of electromagnetic torque, and then reduces the system performance and even leads to system instability. To solve this problem, a robust predictive torque control of induction motor based on the discrete hybrid prediction model is proposed. The traditional open-loop model for prediction of stator current is abandoned. An equation for prediction of stator current of the closed-loop model is established, and its stability and parameter design are studied. For prediction of stator flux vector, the traditional voltage model prediction method is abandoned, and a discrete hybrid prediction model integrating the voltage model and the current model is established. The proportional-integral regulator is used to adjust the smooth switching from the voltage model to the current model of the discrete hybrid prediction model. The electromagnetic torque is calculated by the current and the stator flux, which are predicted by the closed-loop model. The proposed algorithm is verified on a 2.2 kW induction motor experimental platform, and the experimental results show the effectiveness of the algorithm.