Using Rough set and RBF Nnetwork on Fault Diagnosis
Li Hang · Equipment Manufacturing Technology · 2010
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.