粗集—RBF神经网络在故障诊断中的应用

Using Rough set and RBF Nnetwork on Fault Diagnosis

将粗糙集—RBF神经网络的故障诊断方法应用于飞机燃油系统的故障诊断,该方法是把原始数据进行属性约简,得到最小条件属性集后,输入后置RBF神经网络进行信息处理,不但能有效地简化神经网络的结构和规模,也提高了故障诊断效率。

Using rough set and RBF network on fault diagnosis of fuel system of plane.The method eliminates unnecessary attributes from the decisiontable,after the minimum fault feature subset is selected,these selected samples will be sent to RBF neural network for diagnosis.The result of simulation indicates that this method can reduce the needed training samples and simplify the neural network structureand size.The efficiency of fault diagnosis has been greatly improved by using the system.