Uncertainty quantification of demand response datasets using peridynamics

This paper investigates curve-fitting methods to explore uncertainty quantification approaches for demand fluctuations in a power grid. Specifically, we use data sets from the Pennsylvania-New Jersey-Maryland (PJM) Interconnection, where we employ two stages of fitting. In stage one, a trigonometric based fit for historical data sets is carried out. In stage two, load forecasting for day-ahead market is developed using peridynamics under uncertainty conditions. The preliminary results from this two-stage approach were promising for resource planning and energy management tasks. This work will allow system operators to capture both stable and non-stable operation early enough to perform planning tasks related to the microgrid.