基于神经网络的现代感应电机自适应L2鲁棒控制

提出一种用于现代感应电机调速的自适应L2鲁棒控制方法。采用反步法(backstepping)推导出2个自适应L2鲁棒控制器，这2个控制器1个控制转速和磁链外环，1个控制定子电流内环。考虑了由感应电机定转子电阻、转动惯量和负载转矩的不确定性造成的扰动，采用RBF(radial basis function)神经网络来补偿这些扰动。提出的控制器可以和1个转子磁链观测器联用。应用转子磁链定向模型固有的解耦性质和HJI(Hamilton-Jaccobi-Issacs)不等式，从全局上证明了控制系统的鲁棒性。仿真结果表明，提出的控制方法对感应电机的不确定性有很强的鲁棒性，且具有很高的动态性能。

Adaptive L2 Robust Control of Modern Induction Motors Using Neural Networks

An adaptive L2 robust control method based on neural networks is proposed for modern induction motors. Using backstepping, the adaptive L2 robust controllers are derived. The disturbances generated by the uncertainties of the stator and rotor resistances, rotor inertia and load torque of an induction motor are compensated using radial basis function neural networks. The proposed controllers are combined with a rotor flux observer. The robustness of the whole control system is proved using the decoupling property of rotor flux oriented model of induction motors and HJI (Hamilton-Jaccobi-Issacs) inequality. The simulation results indicate that the proposed controllers are robust to the considered uncertainties of the induction motor and has high dynamic performance.