Research And Implementation Of The Grid Intelligent Data Analysis Based On Data Mining

In general, data mining is an advanced technology for data analysis,and it focuses on analyzing and understanding data and revealing the essence knowledge and information hidden in some large data sets. In other words, data mining trends to find the useful knowledge and information from some large data sets with the noise information. The patterns of data mining include: classification, clustering, time series, association, sequence, etc. The intelligent analysis of power grid data is to employ different algorithms from the data mining field to analyze intelligently the faults about the electric power equipments, the daily-report data, and the implementary data. More detailedly, the intelligent analysis for electric grid data is based on two steps: first, according to the features from the electric grid implementation and the analyzed factors, it extracts and analyzes some related data from the initial data, and then stores the related data into data storehouse; second, by applying some data mining algorithms, we can obtain some useful knowledge, which plays a theoretical foundation for the security about the electric grid implementation.The intelligent analysis for electric grid data includes four parts, such as: data ETL, knowledge mining, data dynamic update, and data visualization. In general, data ETL implements the cleanout, organization, and loading for data, and it can effectively improve data quality, leading to further boost the performance of data mining algorithms. The knowledge mining plays a significant role in the intelligent analysis for electric grid data, and it can obtain the useful knowledge by applying the data mining algorithms to the data sets from the data storehouse. The data dynamic update focuses on implementing the dynamic knowledge mining on the basis of the variety about electric grid data. The data visualization trends to visualize the finial results.In this thesis, firstly, I give a brief review about the background of data mining (such as some basic concepts, models, etc), and further present the design idea and system configuration about the intelligent analysis for electric grid data. Secondly, I detailedly address the system based on the significant data mining technologies such as data ETL, conjunction rule, time series, data dynamic update, and further I also present some appealing results on the basis of implementing the data mining algorithms via software. Finally, some conclusions are made, and some research directions are also addressed.