A MATLAB based occupant driven dynamic model for predicting residential power demand

Abstract

This paper presents a MATLAB based dynamic model for predicting residential power demand. Markov chain based occupant behavior models developed using data gathered by the U.S. Census Bureau in the American Time Use Survey (ATUS) are used in conjunction with models of the most common residential loads to predict residential power demand on a one-second time scale. First, the methods utilized for the modeling of each residential load are presented. Next, an explanation of how these load models are combined with occupant behavior models to predict residential power demand is given. Simulation results showing the overall contribution of each load to the overall residential sector power demand are shown for both winter and summer cases. Finally, future work will involve the use of this high-resolution dynamic residential model to estimate the potential for demand response from residential loads.