Fault Diagnosis Based on Rough Set and Neural Network Ensemble
Guo Xiao-hui, Xiaoping Ma · Control Engineering of China · 2007
An intelligent fault diagnosis model using rough set and artificial neural network ensemble is developed,considering that neural networks are best for solving nonlinear problems while rough set is good for data reduction.Basis on data acquisition and pretreatment,the original fault diagnosis samples are discretized by using rough set theory.According to the decision attribute positive region of condition attribute(s), the minimum fault feature subset is selected,and the neural network structure is determinted.The networks are trained to reflect the mapping between inputs and outputs,and network ensemble is used to realize the fault diagnosis.An example shows the effectiveness of the method.