Indoor environments should meet the needs of the occupants and enhance their comfort, health and productivity. Efforts to reduce energy consumption often lead to decreased satisfaction for building occupants. These modifications to the environment often cause occupants to change their behavior to improve their personal comfort, often resulting in additional energy use and often cancelling any intended energy improvements.
This study examines how occupant behavior can be more accurately predicted based upon demographics and comfort profiles. Surveys and continuous energy monitoring results provide an in-depth understanding of the indoor environment preferences of the occupants and their energy consumption habits. The data are collected for two case study buildings in Pennsylvania, and two in Doha, Qatar and will be modeled into a machine learning algorithm to forecast occupant comfort desires ad behaviors in a space. A simulation platform is being developed that can accept occupant behavior and preferences as inputs and produce corresponding energy consumption behavior data to help better forecast the user impacts for different design decisions.