基于模糊神经网络的异步电机速度控制系统

针对矢量控制异步电动机，利用模糊神经网络控制理论．提出了一种教师值指导和隶属函数自调整的模糊神经网络控制器。详细描述了模糊神经网络控制器和教师值控制器的设计过程，采用了隶属函数参数在线调节的办法，推导了隶属函数参数调整的公式和算法，研究了以转矩电流Iqs为目标函数和系统输出转速ω为目标函数的情况下的系统转速响应，分析了系统转动惯量J、粘滞摩擦系数B和转矩系数Kt变化情况下系统的性能，仿真结果表明该控制器具有速度响应快、鲁棒性较强等优点。

INDUCTION MOTOR SPEED CONTROL SYSTEM BASED ON FUZZY NEURAL NETWORK

A fuzzy neural network controller with the teaching controller guidance and parameter regulations for vector-controlled induction motor is proposed. The design procedures of the fuzzy neural controller and the teaching controller are described. The parameters of the membership function are regulated by an on-line learning algorithm. The speed responses of the system under the condition, where the target functions are chosen as Iqs and ω, are analyzed. The system responses with the variant of parameter moment of inertial J, viscous coefficients B and torque constant Kt are also analyzed. Simulation results show that the control scheme and the controller have the advantages of rapid speed response and good robustness.