Research And Implementation Of Energy System Which Based On Data Mining Methods

Nowadays our people live in an era that surrounded by information. People’s life would be out of control without information. As time goes on, the data is growing extremely intense. To the data center, that means it would need more computer and more and more data center would be built in various places. It is estimated that the data center’s power consumption almost accounted about1.5%of the total global energy consumption in annual, equally the sum of26power plants generate electricity a whole year, and this number will continue to grow in the future. If we do not use any management to control the power consumption,the energy shortage crisis may appear, our life may be affected by that.In order to reduce the energy waste of the data center, we need research and find the factors which affect the energy consumption of the data center. Use reasonable improvement measures to reduce the overall energy consumption of the data center.These factors include environmental factors, equipment factors and some other factors. In this paper, we mainly analyze the energy consumption data of data center and try to find some methods to reduce the energy consumption, the work includes:1. The energy consumption of internal equipment of data center will be analysis through data mining clustering algorithm. The abnormal equipment will be detected through the difference between the energy consumption of different equipment. As a result, the performance could be improved.2. The historical data will be classified through data mining classification and prediction algorithm. At the same time, the future trend will be predicted. A classification and prediction algorithm of data center will be proposed to control the internal equipment for energy saving.3. An energy consumption analysis system will be established for the practical application of the algorithm mentioned above. Part of the system has been built to provide a platform for clustering, classification and prediction. At the same time, the system will enhance the human-computer interaction.