Net ZEB Design Processes and Tools - Modeling, Design, and Optimization of Net-Zero Energy Buildings - page 217

Stochastic Occupant Modeling
A newer approach to incorporating the effects of occupants on energy use
is to use stochastic (random) models within BPS (Bourgeois, Reinhart, and
Macdonald, 2006; Bradley, 2003; Gunay, O'Brien, and
Beausoleil-Morrison, 2013; Hoes et al. , 2009; Nicol, 2001; Rijal et al. ,
2007). Numerous stochastic occupant models have been developed based
on field studies and experiments to predict the probability of adaptive
occupant actions (e.g., window opening, light switching, blind control)
based on one or more environmental variables (e.g., indoor air
temperature). The majority of these models are represented as
single-variable logistic functions (e.g., Figure 4.19 ) . However, some
stochastic models are intended for specific times of day (e.g., arrival in the
office) when behavior has been observed to be significantly different.
Fig. 4.19 Graphical representation of an example of a stochastic occupant
model for manual window shade control
Stochastic occupant models provide some randomness with regard to
occupant behavior and yield two major benefits: (1) identify peak loads and
(2) prevent designers from optimizing a building to perform around a rigid
set of unlikely occupant-related patterns. More work is required to expand
the field of knowledge and generalize these models to a greater number
of building types, technologies, and climates. A comprehensive review of