An approach of fault diagnosis in nonlinear circuits

Considering the problem of parametric fault diagnosis of nonlinear circuits, an approach based on the wavelet filtering technology and principal components analysis (PCA) is proposed. Firstly, the output signals filtered by DB2 wavelet packet are used to detect the faults. Then, the energetic values in eight subbands are calculated as the principal components. By comparison with the eigenvalues of the energetic relative matrices, different states of the faulty circuits are identified and the faults are located. This provides a novel way to realize fault diagnosis for nonlinear circuits.