Reduction of linearized power systems for the study of interarea oscillations

This paper describes the application of a model reduction method to obtain linear power system models for control system design. The reduction method is based on the computation of an eigensubspace of the product of the observability and controllability Gramians. The technique avoids illconditioned balancing transformations. The method is applied to coherency reduced power systems to generate smaller systems that are suitable for control design applications. These reduced systems retain the interarea modes and their input/output characteristics are similar to those of the coherent models. A 376-state and a 184-state power system models are used to illustrate the method. For the 184-state model, a control design verifies that a 45-state reduced model is a valid design model. For systems with very large sets of state variables, a coherency input-output approach is proposed. In this approach, the reduction method is applied to individual coherent areas