REVIEW OF FAULT IDENTIFICATION USING OF NN BASED GA PARAMETERS IN ANALOG CIRCUITS

Abstract

Fault analysis in analog circuits has been a subject of research for the last few decades because of the complexity in identifies defect models. This paper proposes a fault identify an approach for analog circuits using hybrid evolutionary techniques and neural networks. A neural network is used because of its good robustness and adaptability and a genetic algorithm is used as an evolutionary technique for optimizing and learning neural networks. The proposed method is validated through a state-variable filter circuit and takes all possible parametric variations for faulty and non-faulty conditions and experimental results are presented to show that the comparison of the hybrid scheme neural network method is more efficient.