AGGREGATION PROPERTIES OF LINEARIZED 2-TIME-SCALE POWER NETWORKS

In the slow coherency approach, the slow coherent areas of a power system are aggregated to form a slow subsystem to model the dynamics of the interchange between the areas. This paper addresses the aggregation process associated with the slow coherency approach that uses the singular perturbation technique to correct for the internal connections within the coherent areas to generate improved slow aggregates. Three aggregation properties are discussed: 1) the static load buses can be retained in the aggregation process; 2) the aggregation can be performed on a per-coherent-area basis; and 3) the aggregation process can be decomposed into a sequence of aggregating two coherent machines at each step. These properties allow practical algorithms to be designed for applying the slow coherency approach to large power systems.