A Structure-Preserving Approach to Power System Normal Form Analysis

Large, sparse power system models arise naturally as dynamic models of a wide range of power system applications. In this paper, a structure-preserving approach to the approximate analysis of power system models described by differential-algebraic equations is explored. Using singular perturbation theory, a fast dynamics is added to the algebraic constraints to express the DAE model of the power system in an explicit state space form that can be used for nonlinear analysis of large-scale systems. The theory of normal forms is then employed to construct an analytic approximation to system behavior which retains the physical significance of network variables. By using this new technique, the contribution of the network states to the inter-area oscillations is determined and insight into complex nonlinear behavior is provided. The proposed methodology is demonstrated on a realistic 16- machine, 68-bus dynamic equivalent of the New England test system. Preliminary results show that this approach compares favorably with numerical simulations using full system models, and has the potential to be applied to realistic power system representations.