A network-topology-based load flow for radial distribution networks with composite and exponential load

Abstract

A network-topology-based method to solve the load-flow problem of radial distribution networks is reported in this paper. The proposed technique, based on network graphical information allows power flow equations formulation in matrix form to satisfy the need of distribution automation. The technique only requires tabulation of the input information for line data in such way that the receiving end node must be in an ascending order. A directed graph, of a radial network represented by a nodes-by-nodes sparse matrix allows detection of the path of power flow from the reference node to the leaf end. Traversing the directed graph in depth-first search form, the power flow paths (downstream nodes for each node including the node itself) are detected. A BN connection matrix is constructed based on the discovered paths. The lengths of the discovered paths explore the number of downstream nodes from each node including the node itself. The bus-injection to branch-current (BIBC) is built by assigning unity to the nodes of the discovered paths. The proposed method (PM) also allows dynamic building of the two matrices: BIBC and branch-current to bus-voltage (BCBV) matrix, used to find out the load flow solution. Reconstruction of these matrices takes place automatically by just changing two elements in the sparse matrix S reflecting changes in network configuration. The PM is compared with the other methods to demonstrate its effectiveness. The convergence ability of method is also evaluated for different types of load-modeling, different tolerance values, loading conditions and r/x ratios. The results of the applications of the proposed methodology to a set of networks taken from the literature on the topic along with conclusion are presented.