Analog Circuit Fault Fusion Diagnosis Method Based on Support Vector Machine

Abstract

Lack of fault samples and statistic characteristic of artificial neural network, which restrict its more development and application in fault diagnosis. Support Vector Machine (SVM) is a machine – learning algorithm based on structural risk minimization principle, it has the capability of solving commendably learning problem with few samples. Now, analog circuit fault diagnosis on SVM mainly by single information, the diagnosis result is uncertain. Multi-source information fusion technology is introduced to diagnose analog circuits by integrating multi-source information. An analog circuit fault fusion diagnosis method based on SVM is proposed, binary classification algorithm of SVM is introduced and multi-fault SVM classifiers are developed in the paper. An analog circuit’s multi-fault are classified, the results show that the proposed method has many advantages, such as simple algorithm, quick fault class, good classification ability, approving diagnosis purpose with few samples and so on.