Decomposition techniques in power-system network load-flow analysis using the nodal-impedance matrix

Partitioning techniques for the nodal-impedance methods of the power-system network load-flow solution are derived and classified. The techniques include normal matrix inversion using partitioning and successive inverse modification, the topological or diakoptic and the algebraic partitioning principles. The paper shows how, without reference to network graph theory, these principles follow from simple rearrangement of the matrix iterative procedures.