notesum.ai

Published at November 1

Combining Physics-based and Data-driven Modeling for Building Energy Systems

eess.SY

cs.AI

cs.LG

cs.SY

Released Date: November 1, 2024

Authors: Leandro Von Krannichfeldt¹, Kristina Orehounig², Olga Fink

Aff.: ¹Swiss Federal Institute of Technology Lausanne (EPFL), Route Cantonale, Lausanne, 1015, Switzerland; ²Swiss Federal Laboratories for Materials Science and Technology (Empa), Überlandstrasse 126, Dübendorf, 8600, Switzerland

Arxiv: http://arxiv.org/abs/2411.01055v1

Feature group	Feature variables
Datetime	Season, week (weekday/weekend), daytime (morning/afternoon/evening/night)
Weather	Drybulb & dewpoint temperature, diffuse & diffuse solar radiation, rel. humidity, wind direction & speed
Building	Total cooling & heating mass flows, network temperature, air-conditioning mode (cool/heat)
Room	Mass flow, temperature setpoint, occupancy, window position (closed/open), blinds position (up/down)