| With the rapid development of modern information technology,lots of data of many fields of social life are accumulated in the production and living.It is hoped that through the data analysis and study,and found the realistic needs,which are potential in the knowledge and rule of these data,which make the rise of Datamining subject and the development of Data mining technology.As complex comprehensive field in which multidisciplinary cross,Data mining involves Database,Statistics,Machine learning,High-performance computing,Pattern recognition,Neural network and Data visualization and other disciplines,which is more concerned hot-spot field in the modern computer science.As an important method of Data mining,Decision tree algorithm is essentially the process of generating a decision tree by scientifically analysis to the existing data collection,and through the machine studying.The decision tree algorithm has been widely applied in the field of automatic knowledge acquisition.The most typical decision tree classifier learning algorithm is the decision tree C4.5algorithm,which uses the top-down divide-and-rule strategies and makes use of the concept of information gain to select splitting attributes,and which can make sure to generate a resoluble decision tree.The decision tree C4.5 algorithm is simple and efficient,and the generated knowledge is easy to be understood.Our country is a big country to consume electrical energy.As the greatly improvement of the materialization of people’s life,the demand of the entire society for the electricity also presents the explosive growth.Whenever climate extremes,the status of the demand for electricity in short supply became normalcy.In order to realize the stable operation of the power grid,and to ensure the orderly operation of social order,when the electricity supply and demand are in a critical point,the electric power enterprises often use the method of "orderly electricity" to ensure electricity agency reasonable,and make use of removing those low cost-performance power units lightenin to ensure the power supply of other power units in the society.Power,as a kind of economy society’s commodity,has its irreplaceable value attributes.Due to electricity’s indispensability,the public has become increasingly dependent on the commodity,which requires that the power supply enterprises could supply better,more stable and more cheaper electrical commodities.As the transition link,the transformer substation is the key node to connect the power generation and sell electricity,and so,toimprove the reliability and intelligence of transformer substation is the most urgent task for the development of our country’s power grid.Through the computer technology to realize the intensive,intelligent and information of the electric power industry,we must improve the level of enterprise production and management,constantly reduce the production cost of enterprises,and further improve the labor productivity of enterprises.This paper,based on the classical decision tree algorithm,hopes to solve the practical problems in power production by this technology,and give the data guidance for the future production by analyzing the accumulated historical data in the past production. |
