RILin tietoisku / Bulletin: Modelling individual - Building energy use and indoor health exposures for urban areas using machine learning


Welcome to follow the presentation "Modelling individual - Building energy use and indoor health exposures for urban areas using machine learning". The presentation will be held via remote access.

This talk will describe ongoing research at the UCL Institute for Environmental Design and Engineering in London, modelling the relationship between building energy consumption and the health and comfort of building occupants at the population-level.

At the basis of the model is the building physics tool EnergyPlus, which can dynamically estimate indoor environmental conditions and energy use. A python-based tool is able to batch-generate an unlimited number of EnergyPlus models based on inputs from either building stock models, or randomly selected from defined distributions. A neural network metamodel is then generated, linking the EnergyPlus inputs to the modelled outputs, allowing for the rapid calculation of indoor conditions and energy consumption at the urban or national level. The tool can be used to rapidly calculate building performance and health impacts under a number of different energy efficiency and climate change adaptation policies, and climate scenarios.

The presenter of the bulletin is Jonathon Taylor, a lecturer in Urban Physics and senior researcher at the Institute for Environmental Design and Engineering, Bartlett School of Energy, Environment and Resources, University College London (UCL), UK. He holds a BSc in Biological Sciences from Simon Fraser University, Canada, and a BEng in Geomatic Engineering and PhD in Environmental Building Science from UCL. His research interests are in building physics, spatial data analysis, and health impact modelling. He currently lives in Helsinki with his family, returning to London to teach or attend meetings when necessary.

When you sign up to this event, you get an email where is the Skype-link for the presentation; using this link you are able to follow the presentation via your own computer.