Application And Research For The Grid Caculation On Power System

Grid computing is a faculity that integrates distributed resources together in different places. It integrated computing, data storing, wide network, instruments and equipments into a whole. People can use those resources easily in this faculity which contains many kinds of useful resources. Grid computing depends on network and it serves the network.JXTA-P2P is a set of protocol which is developed by Sun Microsystem. These protocols are used for building Peer-to-Peer communication network. They give us a method for contracting P2P network. Software developers can design the P2P application in Java using JXTA. So JXTA has many kinds of Java's advantages.When human maintain power system to make it run stably, they need to have a better method for the fast state estimation of power system. It is the cornerstone of the run and control of power system. Classic technique is the Newton-Raphson iterative. But when the system grows much bigger, the computing is much more hugeness. And it is very difficult to get the result in a short time under the current hardware and software situation.This thesis which depends on futher learning grid computing framework introduces building P2P network using JXTA, designing the communication frame. It designs the Peer pool to maintain those Peer resources. At last it is applied to the huge matrix computing. It disassembles the huge task into smaller ones according to the special features of the power system such as a great deal of data and a lot of computing. This thesis gives the implemented grid programming in JXTA and distributes the power flow calculation task into smaller ones. Those smaller tasks are solved in grid. It considered the expansibility to more function when implementing the grid. It uses IEEE14 and IEEE30 nodes data to do the testing in grid on multi-computers. At last, it gives the result which is solved on the grid, and the result is compared with the one which is solved in the classic method on single computer.Through comparison of the result which is solved in grid and the one in classic method, it will get a good speed for the flow calculation of the power system because it combines the grid computing and task disassembling algorithm. If grid computing is used in the huge matrix calculation in the power system, it will increase the computing speed through test. So grid computing will have a good foreground in power system.