Development of bilinear power system representations for small signal stability analysis

In this paper, a new analysis technique for predicting and characterizing nonlinear behavior of stressed power networks is proposed. Making use of an analytical approximation of the system nonlinear model via the use of a truncated Carleman linearization technique, a bilinear state-space model of the power system is developed in which the second and higher order nonlinear terms are explicitly incorporated in the series expansion representation of the system model.The proposed framework enables the study of the dynamic behavior of nonlinear systems, both analytically and numerically, and can be used to represent a wide class of non-linear systems and oscillatory processes.Analytical criteria are developed based on the structural properties of the bilinear state-space model matrices, which predict the existence and stability character of modal interaction in terms of the eigenstructure of the linear system representation. The properties and behavior of the bilinear model are then investigated, and a number of useful results are derived. The present method is quite general and extends readily to higher-dimensional systems.A simplified 2-area, 4-machine system is used to illustrate the proposed procedure. Detailed nonlinear time-domain simulations are conducted to identify the strength of nonlinear behavior arising from interaction of the fundamental modes of oscillation, as well as to check the validity of the analysis.