SVM binary tree multi-classification method based on IBQPSO and its application in analog circuits fault diagnosis

Abstract

Aiming at the problems of traditional classification methods, a SVM binary tree multi-classification method based on Improved Binary-coded quantum particle swarm optimization (IBQPSO) is proposed in this paper. First, IBQPSO algorithm and its steps are presented and multi-classification SVM based on binary tree is discussed. Then, the presented method and procedures are analyzed. Finally, taking the biquadratic filter circuit as an example to do the simulation experiment, the result shows that the proposed method is efficient.