Efficient heuristic partitioning algorithm for parallel processing of large power systems network equations

An efficient heuristic algorithm for automatic partitioning of a power system network is presented. The algorithm was designed for use in conjunction with a diakoptics-based coarse grain parallel method for solving large, sparse linear systems arising from network analysis. It exploits the techniques of factorisation path graph partitioning and equivalent post-ordering. Test results for systems up to 811 busbars and 1476 transmission lines are included for comparison purposes. This indicates that the method is able to divide a power system network model into a number of equal sized subnetworks in order to optimise the use of parallel computer systems for network analysis while the resultant number of fill-in elements is kept to a minimum.