Application Of Decision Tree Algorithm In Substation Fault Diagnosis

At present,the electric power industry is developing rapidly and the number of substations is increasing,the power system structure is gradually complex,and the monitoring system collects a large amount of data information containing a lot of noise in real time,which interferes with the extraction of effective data information and fault diagnosis of the power system.In order to avoid the above,we use data mining technology to extract and analyze valuable operational diagnostic information from massive data.Therefore,it can diagnose substation power equipment quickly and accurately,shorten the operation time of faulty equipment and improve the safety and stability of power system.This paper first describes the concept of data mining and common algorithms,as well as the application of data mining technology in power fault diagnosis.Typical warning signals are selected for analysis and classified and combined.Then the model of decision tree algorithm is constructed,and several classical decision tree algorithms are elaborated,as well as the comparative analysis of each attribute.Then,the decision tree algorithm is improved from three aspects of optimizing sample size,optimizing test attributes and discretization.Sample data was selected to verify the algorithm model,and the results showed that the classification accuracy and modeling speed of the optimized decision tree algorithm were better than the classical algorithm.Finally,a power fault diagnosis system is constructed by combining the data mining technology with the power diagnosis expert system,and the improved decision tree algorithm is applied to the system.The power diagnosis expert system based on decision tree optimization algorithm is composed of inference engine,interpreter,decision tree algorithm and graphic interface.The real-time data collected by the monitoring background is applied to the diagnosis model,and the typical fault information is adopted for verification.The results show that this system can quickly and accurately diagnose the power equipment and its fault causes.Its realization process is simple and the applicability is high.It also has the very good use value and the significance.This papers has 17 figures,9 tables and 51 references.