Application of rough set and support vector machine in fault diagnosis of power electronic circuit

Fault diagnosis of power electronic circuits is very important in power system. Fault elements of power system are found quickly and correctly by fault diagnosis of power electronic circuit. Fault diagnosis method of power electronic circuit based on rough set and support vector machine is presented, where support vector machine (SVM) is a machine learning method to solve a binary classification problem in a supervised manner, rough set is used to simplify redundant attribute. A certain power electronic circuit is used to testify the diagnostic ability of rough set and support vector machine. The comparison results among RS-SVM, SVM and BP indicate that RS-SVM has higher diagnostic accuracy than SVM, BP classifiers.