Fault diagnosis of power transformer based on modified decision directed acyclic graph SVM

作者：Zhou Zhicheng; Ge Le; Tao Fengbo; Li Jiansheng; Yang Chengshun

来源：ICIC Express Letters, Part B: Applications , 2016, 7(8): 1843-1849.

• Decision directed acyclic graphs
• Fault diagnostics
• Multiclass SVM
• Node Optimum
• Transformer fault diagnosis

The method for fault diagnosis of power transformer based on the modified decision directed acyclic graph (DDAG) support vector machine (SVM) is proposed to enhance the accurate rate of the fault diagnosis of power transformer. DDAG is an extending scheme with superior performance, but its decision mainly depends on the sequence of nodes which is arbitrarily selected. So, in this paper, a modified DDAG based on node optimum algorithm is developed to improve the the accuracy of diagnosis, firstly. Then, multi-class SVM extended by the proposed method is applied to transformer fault diagnosis, and the fault diagnostic process based on modified DDAG-SVM is developed in detail. The validity and correctness of the proposed method are verified through experiments.

• 出版日期2016
• 单位南京航空航天大学; 南京工程学院