The Design And Realization Of The Functional Components Of The Power System Grid Computing Platform

With the development of electric power industry, power system becomes a large scale system dispersed in a wide area. This brings two fundamental problems: 1) How to manage the information and resource efficiently which are dispersed in a large system;2) How to analyze the large-scale system in a real-time and timely manner. Although the past distributed technology can solve these problems to some degree, it is not able to satisfy the operation and control of modern power system. Hence, it is necessary to introduce the novel distributed technology---Grid Computing, to the application of electric power engineering, in order to explore new method which will make control and decision-making on large electric network more efficient.In order to accomplish this purpose, the power system automation laboratory at UESTC developed PSGridâ—‹R (Power System Grid) based on Globus3.0. This thesis describes the design and implementation of the functional components in this platform. These components are categorized into two types: One is to serve the management of distributed information, including the virtual database querying component and the dynamic data supervising component ; Another is to serve the distributed parallel computing in power system, including the matrix fraction component and the parallel computing component.In the thesis, the fundamental knowledge about Grid Computing is introduced, followed by the description of the design and implementation of these functional components . The steps and methods for deploying the components as a grid service are also presented. At last, the components are tested on PSGridâ—‹R . The results show that all the functional components meet their demand of design. And performances of the parallel computing component are evaluated by test, in which two Jacobi matrix equations of power flow calculation with different dimension are selected as the test case. The results are compared with that of serial computing and MPI (Message Passing Interface). It is shown that the components proposed in this thesis not only has better performance of acceleration, but also can support parallel computing for power system whose data is distributed in a wide-area network.