Occupancy Profile Variation Analyzed through Generative Modelling to Control Building Energy Behavior

Nowadays, building energy models use parametric analyses to optimize design strategies considering multiple variables.
Integrated dynamic models combining design tool and visual programming language (VPL) and simulation tools to calculate
building performance with BIM tool for the whole-building energy simulation have been adopted in the recent studies. Through
these tools, it is possible to identify parametric systems, which become a “genome”, where a rapid comparison of different
alternatives is possible through fitness criteria defined by design goals. The aim of the paper is to use this concept and the
suitable parametric tools such as Grasshopper for Rhinoceros to handle variable hypotheses on users’ occupancy that influence
building energy performance. The paper focuses on occupancy variability applying the methodology to a university building
located in northern Italy in the University of Brescia Campus to evaluate how generative modelling can represent an adequate
approach to energy simulation of occupant behaviour. Sensors are now monitoring the real occupancy trend of the case study
building and different scenarios defined in the parametric model could be compared to the real weekly. Using parametric tool and
GA (Genetic Algorithms) can be analysed hundreds of occupancy patterns in order to better understand the influence of the
occupancy on the building energy use and at the same time evaluate different strategies to save energy.