Abstract by Jacob Nuttall
Physics and Astronomy
Power Systems: Formalizing the Method of Constructing Dynamical Equivalents for the Grid
The power grid is a complex network of components that dynamically responds to varying loads and inputs and whose stability is perturbed by short-circuits and other abrupt disturbances. Effective control of the grid requires models that can make accurate predictions in real-time. One of the challenges to predictive modeling is the need to estimate many unknown parameters by fitting data taken from transient disturbances. However, most parameter combinations are unindentifiable, i.e., unconstrained, with available data, so it is common practice to compress large areas of the grid into "dynamical equivalents": simpler circuits with a few parameters that are tuned to match observed data. We attempt to formalize the construction of these dynamical equivalents through constructing reduced versions of standard power systems models with respect to two different areas of observation.