| Recently the power system industry has witnessed an increase in system automation affecting all aspects of operation and requiring knowledge of system states that are typically provided by system state estimators. Traditionally, if the state estimation problem is solvable, the power system is considered observable. The state estimation problem considers a static model of the power system (static state estimation), represented mathematically by a nonlinear algebraic model, derived from the system measurements, and ignores the nonlinear dynamics of the system. This research presents a case for dynamic state estimation and nonlinear observer-based controllers that are required for advanced system control, by first addressing the problem of qualifying and quantifying system observability for dynamic power systems, and incorporating the ability to track system observability along trajectories of the dynamic system states in time. The formulation is derived from a Differential Algebraic Equation (DAE) model of the power system, accounting for both the nonlinear dynamics and algebraic constraints of the system. Computer simulations outlining the importance and benefits of the new method are presented, emphasizing on three main aspects of the formulation: (a) the effects of the power system models, (b) the effects of the measurement system, and (c) the dimensionality of the observability problem. To illustrate the characteristics of the observability formulation, results were obtained using typical power systems. |
