Soft Fault Diagnosis of Power Electronic Circuits Based on GSABO-VMD-CNN-BiLSTM Model
FAN Wanbei
JANG Yuanyuan
Abstract：To address the low diagnostic accuracy issue caused by insufficient signal features and noise in the soft fault diagnosis of traditional power electronic circuits,a DC/DC golden sine adaptive backpropagation optimization-variational mode decomposition-convolutional neural network-bidirectional long short-term memory(GSABO-VMD-CNN-BiLSTM)model was proposed for soft fault diagnosis of power electronic circuits.Firstly,the GSABO algorithm was applied to optimize the VMD parameters in order to solve the problems of mode aliasing and endpoint effects.Secondly,the minimum envelope entropy and minimum arrangement entropy were combined to construct a composite fitness function,and the wavelet threshold function was utilized for denoising to improve the data quality.Finally,the time-domain features were extracted and input into the CNN-BiLSTM model to complete the fault diagnosis.The model was experimentally verified by the 150 W Boost circuit,and the results showed that the accuracy of the model reached 99.58%.And under different signal-to-noise ratios,the model performed well in terms of accuracy,recall,and other indicators.The model could be effectively used for soft fault diagnosis of power electronic circuits.
Keywords：DC/DC circuitsvariational mode decompositionwavelet thresholdcomposite fitness functionCNN-BiLSTM
Publication Date：2025-06-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 259-265 )