A Power Grid Fault Diagnosis Method based on Ensemble Decision Tree

Abstract

The reason for the failure of the grid is complicated. On the one hand, because the majority of the information comes from the grid signal whose description dimension is higher; on the other hand, it is a low probability that large number or types of fault problems occur in one area. To solve this problem, this paper proposed a power grid fault diagnosis method based on ensemble decision tree (PGFD-EDT). This algorithm uses attribute selection mechanism to divide a large number of power signal attributes for multiple sets of attribute subsets, each subset is trained individually for a decision tree, and multiple decision tree models vote together to determine the grid failure. Experiments show that the PGFD-EDT algorithm has higher stability and accuracy.