Paper Conference

Proceedings of BSO Conference 2018: Fourth Conference of IBPSA-England


Data-driven Behavioural Modelling for Building Energy Simulation Based on Scripted Profiles

Federica Naspi, Marco Arnesano, Gordon Aird, Birthe Klebow, Francesca Stazi, Marco D’Orazio, Gian Marco Revel

Abstract: Energy simulation software describe users’ behaviours with static profiles, causing significant discrepancies between actual and predicted building performances. A User Behavioural Module (UBM) tool has been developed using a data-driven approach and coupled with the IESVE energy simulation software to stochastically reproduce users’ behaviours. Initial investigation on the required number of simulations showed a deviation on the overall consumptions of about 1% in a range between a single and 20 runs of simulation. The simulated electricity consumptions varied up to 10%, applying the UBM instead of Standard profiles. The tools interoperability provides considerable enhancements to the engine and its outputs.
Pages: 331 - 337