Remote Power Outage and Restoration Fault Tracing Based on Classification Decision Tree-Association Rules
The electricity cost remote control system has become an important link in smart grids,but the power outrage and restoration fault rate of the cost control system has reached 10.64%.At present,only 1/3 of the faults are gradually checked by manual labor based on historical experience,and the rest of the faults cannot be found.In order to realize the fault diagnosis quickly,and promote the construction of smart grids,based on decision tree association rules,this paper proposes a method for tracing power outrage and recovery failures.The decision tree algorithm is used to classify different types of data,and the attribute feature table is established accordingly.The information gain rate of fault features is calculated,and the decision is divided to obtain the data set of fault causes.The association rules are used to calculate the degree of improvement of fault→cause,find out the strong correlation between each fault and cause,establish fault traceability rules,and realize fault traceability.This method can quickly find the cause when the fault occurs and improve the work efficiency,and the effectiveness and efficiency of the proposed method are verified through examples.
the power recovery systemfault traceabilityinformation gain ratedecision treeassociation rules
NETL
NSTL
万方数据
