Analog circuits fault diagnosis based on Adaptive Fuzzy Neural Network

This paper mainly presents an analog circuit fault diagnosis by adaptive fuzzy neural network. Combines fuzzy theory with BPNN(back propagation neural network), an integrated self-adaptive NN is developed based on the Takagi-Sugeno fuzzy system. The training of network weights and optimization of membership functions are conducted employing hybrid algorithms. Finally, single electrical source complementary symmetry power amplifying circuit is illustrated. The feasibility and validity of the method are validated by simulation testing.