模糊神经网络技术在电机调速中的应用

Application of Fuzzy Neural Network to Variable Speed Motors

基于模糊神经网络技术提出了一种感应电机定子电阻在线估测的有效方法．在对广泛选择的样本进行学习后，优化了控制规则、各语言变量的隶属函数及每条规则的输出函数，在线估测结果与实验结果吻合良好．为进一步估算直接转矩控制或矢量控制系统中电机的磁通提供了可靠的保证．

An effective method, based on fuzzy neural network, is proposed to estimate the real time stator resistance of the induction motor. The control rules and the membership function of every subset, together with the output functions, are all optimized by learning the samples, evenly selected from the definition domain of each variable. The estimated results agree quite well with those gained from the experiments. This provides a reliable guarantee for further estimating the magnetic flux of the motor in the direct torque control system or the vector control system.