Fault Diagnosis for Transformer Based on Random Forests
Tao Dongq · Dianzi qijian · 2015
To study the fault diagnosis of power system,safe and stable operation of the transformer is of great significance. The content of the various characteristics of gases dissolved in transformer oil setting-up for the decisionmaking analysis is the important methods for the detection of transformer. In view of the single decision tree classification,its effect is poor,poor in anti-interference ability. The method of using random forests is put forward for fault classification,and a combined classifier model is set up by the combination of classifiers,high precision,strong stability and not fitted,so that the occurrence of fault can be more timely and effectively diagnosed,the normal operation of the transformer can be ensured.