Neural Network Ensemble Approach in Analog Circuit Fault Diagnosis

Abstract

Neural Network (NN) based analog circuit fault diagnosis approach is the widely used strategy in present. In this paper, a NN-ensemble-based strategy has been introduced into the field of analog circuit fault diagnosis. Efficient, accurate and different fault features sets are obtained by resampling the original feature sets with Bagging algorithm in order to train individual RBF neural networks as component classifiers simultaneously, then plurality voting strategy is employed to isolate the actual faults of analog Circuit Under Test (CUT). Experimental results indicate that compared with any of its individual RBF neural network, the NN ensemble is able to effectively improve the generalization ability of the analog circuit fault classifier and increase the fault diagnosis accuracy.KeywordsNeural network ensembleBaggingAnalog circuitFault diagnosis