一种新型径向基函数神经网络在模型辨识方面的应用

A New Radial Basis Function Neural Network Application for Identification Model

采用改进的径向基函数(MRBF)神经网络,利用实测到的发动机飞行试验数据作为学习样本,建立了发动机的辨识模型,并利用这种方法对不同飞行高度发动机的参数进行了辨识。研究结果表明:这种方法具有训练时间短、学习速度快、辨识精度高、实时性好等优点,并可用于在线辨识。

The application for identification model based on the modified radial basis function network was set up by using the measured flight tests data as learning stylebook. The parameters of engine were identified at different flight heights by this method. The results show that this method has the advantage of faster learning rate, higher identifying precision, better real - time ability and easier to bring into effect.