Application of ant colony optimization-SVM in fault diagnosis for rectifier circuit

Failure of rectifier circuit has the characteristics of latency and complexity, which leads to the difficulty to fault diagnosis for rectifier circuit. A new method of optimizing support vector machine (SVM) by using ant colony optimization algorithm is presented to fault diagnosis for rectifier circuit in the paper. The experimental object is provided and the six ACO-SVM classifiers are developed to identify the following seven states of the experimental object. The testing results demonstrate that the ACO-SVM classifier has higher diagnostic accuracy than normal support vector machine and BP neural network.