The Research And Application On The Power System Multi-source Heterogeneous Data Optimization Management Technology

With the accelerated pace of the national power grid, the data of power company show an exponential growth. In the data of Power Company, multi-source heterogeneous data are the main component. In the face of a large number of multi-source heterogeneous data, how to store, call, calculate and manage it effectively has became an urgent problem to be solved. Aiming at this problem, this thesis designs and realizes the data Gan shifting storage system. This system can achieve optimal management of power grid enterprise for its daily data, and through the virtual machine cluster Gan moved to the data management center and data management center will be divided into structured data and non structur ed data. And archive the data to the corresponding memory address automatically.The contents of this thesis include the following points:1. This thesis describes the background and current situation of multi-source heterogeneous data management technolog y to optimize power system, combined with multi-source heterogeneous data management system optimization, and describes the purpose and significance of data migration and storage.2. In this thesis, faced the problem of the existence of unity for the trad itional virtual machine data migration method, using the unity of data migration method which is based on GA optimization algorithm can improves the concurrent data migration method. According to the above method on a virtual machine cluster platform, validating test concurrent data migration methods, experimental results show that concurrent data migration method can enhance the data migration of virtual machines effectively and reduce virtual machine migration operation latency and overhead.Then it will improves the resource utilization of virtual machine eventually.3. To achieve the goals of structured and unstructured data of unified management, we have studied the system which stores procedures and application examples of multi-source heterogeneous da ta, have analyzed the characteristics of the storage and management status of unstructured data, and have analyzed the parallel computing multi-source heterogeneous data management system optimization. It was designed based on applications to optimize mult i-source heterogeneous data management system.4. According to the scheme above, the optimization system of multi-source heterogeneous data is carried out in the network business field. The system has function of data collection, data analysis, data retrieva, etc. Doing test for boot visual machine, checking its state and shut ing down in the trucked visual machine platform, the result of test shows this system had a high feasibility.