Feature Extraction and SVM Parameters Optimization Based on ICQPSO

In order to further study the fault diagnosis of analog circuit and improve the effectiveness of the application of SVM parameter optimization method, a fault diagnosis method of analog circuit based on ICQPSO feature extraction and SVM parameter optimization is proposed by combining the feature processing method with SVM parameter optimization from the perspective of feature extraction. In order to improve the classification accuracy and generalization ability of SVM, ICQPSO is used to optimize the transformation matrix and SVM parameters. The simulation results show that this method is very effective for the construction of high-precision SVM, and it is feasible to apply it to the fault diagnosis of analog circuits.