Research On Application Of Data-mining In Supervision And Management Of Power-grid Safety

Together with the development of economy,China's power grid construction is in a high-speed development phase also.Recently,China has already constructed UHV and EHV transmission lines successively.With the scale continues to expand,the power grid is becoming more complex,and the links among transmission grid regions are getting increasingly closer.This leads to an increase of abnormality information among transmission lines.The frequent occurrence of power system faults disturb the convenience of the power supply regions and the benefits of the grid companies.Diverse and complex faults caused by weather,seemingly unrelated to each other,however,have practical correlations among various attributes of the grid.Using data mining technology in historical grid faults,to automatically classify and study historical grid fault data,will reveal the potential links and apply the analysis results to fault diagnosis and prediction analysis to provide useful guidance for ensuring the stable operation of the transmission grid.Firstly,this paper use the historical fault information of overhead transmission lines as the data source to classify and sort the equipment involved in overhead transmission lines,and study the influences and damages of weather on the normal operating overhead transmission lines.Secondly,the paper introduce data mining technology in detail,including the history of data mining technology,its related decision tree algorithms and association rules,and analyzing respective examples.Through the improved data mining algorithm,the sorted historical data is analyzed and researched.The useful information for fault diagnosis and predictive analysis of the power system is found out.The results are analyzed and referenced.Thirdly,this paper redevelop and redesign the grid safety production supervision and management system in which data mining technology can be used,and system information can be displayed.