Simulation circuit fault diagnosis method on the basis of generalized multi-nuclear support vector machine

Abstract

The present invention provides a simulation circuit fault diagnosis method on the basis of a generalized multi-nuclear support vector machine. The method comprises the following steps: (1) collecting time domain response signals of a simulation circuit, namely collecting output voltage signals of a simulation circuit; (2) performing Wavelet Transform of collected voltage signals, taking energy used for calculating wavelet and coefficients as characteristic parameters, wherein the set of all the characteristic parameters is sample data; (3) applying regularization parameters and trade-off parameters of a PSO optimization generalized multi-nuclear support vector machine based on the sample data, and constructing a fault diagnosis model on the basis of GMKL-SVM; (4) taking the constructed fault diagnosis model on the basis of GMKL-SVM as a classifier, and performing fault diagnosis of the simulation circuit. The classification performance of the GMKL-SVM is better than other classification algorithms, and the method for optimization GMKL-SVM parameters by applying PSO is better than a traditional method for obtaining parameters so as to efficiently detect element faults of a simulation circuit.

Description

technical field [0001] The invention belongs to the fields of machine learning and electronic circuit engineering, and relates to a fault diagnosis method for analog circuits based on generalized multi-core support vector machines. Background technique [0002] Analog circuits are widely used in household appliances, industrial production lines, automobiles, aerospace and other equipment. The failure of analog circuits will cause performance degradation, functional failure, slow response and other electronic failures of the equipment. Correct identification of analog circuit faults is helpful for timely maintenance of the circuit, so it is very necessary to diagnose the faults of analog circuits. [0003] For the fault diagnosis of analog circuits, existing research work has adopted the artificial neural network (ANN) method. However, the ANN method is generally difficult to determine the structure, the algorithm convergence speed is slow, and it is easy to cause overfitting...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the fault diagnosis of analog circuits, the existing research work adopts the Artificial Neural Network (ANN) method. However, the ANN method is generally difficult to determine the structure, the algorithm convergence speed is slow, and it is easy to cause over-fitting problems.

Support vector machine (support vector machine, SVM) is based on the VC dimension theory of statistical learning theory and the principle of minimum structural risk, which can better solve the small sample problem and nonlinear problem in classification. The setting of kernel function is the key of SVM algorithm , generally by using a single-core learning method, which is simple to operate, but it is easy to ignore the useful information in the input samples, and it is difficult to achieve the optimal generalization ability

Images