Abstract
Stochastic person-based activity models play an important role in the prediction of the realistic time-series energy demand for residential buildings. These models generally use input parameters developed based on time-use data. This paper models the parameters for a stochastic person-based activity models that can consider the variability in simulated activities among households. Modelling parameters were modeled by two steps: 1) classifying time-use data for six segments using the basic demographic conditions, 2) developing regression models for each segment considering detailed demographic conditions as explanatory variables. The developed regression models were validated by the Hosmer-Lemeshow goodness-of-fit test. Then, we compared the model performance with the conventional model only using the sample distributions of each six segment. Results showed that the proposed model improves the producibility of the variability among simulated occupants and households. The use of the developed model would contribute to improving the accuracy of residential energy demand models.
