Aiming at the difficulty of analog circuit fault feature extraction, an analog circuit fault feature extraction method based on wavelet packet energy spectrum and independent component analysis is proposed. Firstly, the fault output signal of the circuit is obtained through simulation, the output signal is decomposed and reconstructed by wavelet packet analysis, and the energy of each frequency band is obtained as the fault eigenvalue through the reconstruction coefficient. Then the independent component analysis algorithm is used to optimize the fault eigenvalues, so as to construct the eigenvector reflecting the circuit fault. Finally, the support vector machine is constructed, the fault feature vector is input for training and testing, and the accuracy of circuit fault diagnosis is obtained. Simulation results show that this method can effectively extract the feature parameters that can feature circuit faults, and the diagnosis accuracy can reach 95.7%.