APPLICATION OF PARTITIONING TECHNIQUES FOR DECOMPOSING LARGE-SCALE ELECTRIC-POWER NETWORKS

Two efficient heuristic algorithms for solving cluster problems associ ated with partitioning of power networks are presented in this paper. Both algorithms are divided into two stages. In the first stage, an in itial partition is created based on the electrical distance among syst em buses. The second stage involves interchanging pairs of buses among the various clusters of the initial partition. The first algorithm so lves an optimal k-decomposition problem. In the optimal k-decompositio n, the partitioning of a network into k clusters of buses is performed by maximizing the number of uncut links among clusters. This type of decomposition is known to be equivalent to a 0-1 quadratic programming problem which is approximated by a linear transportation problem. By defining a link weight as the electrical distance between linking buse s, the algorithm clusters strongly-connected buses together while weak ly-connected buses are placed in different clusters. The second algori thm solves the placement problem of n-connected buses in a k-dimension al Euclidean space; such a problem is reduced to finding k eigenvector s of a connectivity matrix, defined as the bus admittance matrix. The node interchange is an iterative heuristic method that can be used to improve an initial partition, such as that obtained by either the k-de composition technique or the eigenvector approach. The method moves on e bus at a time, from one cluster of the initial partition to another, in an attempt to maximize the total electric distance among clusters of the final partition. Applications of the proposed algorithms to bot h small and medium-size power systems, and comparing the results with those obtained in another study, are illustrated in this paper.