Managing the uncertainty of occupant behaviour for building energy evaluation and management
The influence of building occupancy and user behaviour on energy usage has been identified as a source of uncertainty in current understanding of operational buildings, and yet it is rarely directly monitored. Gathering data on the occupancy of buildings in use is essential to improve understanding...
| Main Author: | |
|---|---|
| Format: | Thesis (University of Nottingham only) |
| Language: | English English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | https://eprints.nottingham.ac.uk/51466/ |
| _version_ | 1848798502941884416 |
|---|---|
| author | Naylor, Sophie |
| author_facet | Naylor, Sophie |
| author_sort | Naylor, Sophie |
| building | Nottingham Research Data Repository |
| collection | Online Access |
| description | The influence of building occupancy and user behaviour on energy usage has been identified as a source of uncertainty in current understanding of operational buildings, and yet it is rarely directly monitored. Gathering data on the occupancy of buildings in use is essential to improve understanding of how energy is used relative to the actual energy requirements of building users.
This thesis covers the application of occupancy measurement and processing techniques in order to address the gap in knowledge around the contextual understanding of how occupants’ changing use of a building affects this building’s optimum energy demand in real time. Through targeted studies of running buildings, it was found that typical current occupancy measurement techniques do not provide sufficient context to make energy management decisions. Useable occupancy information must be interpreted from raw data sources to provide benefit: in particular, many slower response systems need information for pre-emptive control to be effective and deliver comfort conditions efficiently, an issue that is highlighted in existing research. Systems utilising novel technologies were developed and tested, targeted at the detection and localisation of occupants’ personal mobile devices, making opportunistic use of the existing hardware carried by most building occupants. It was found that while these systems had the potential for accurate localisation of occupants, this was dependent on personal hardware and physical factors affecting signal strength. Data from these sources was also used alongside environmental data measurements in novel algorithms to combine sensor data into a localised estimation of occupancy rates and to estimate near-future changes in occupancy rate, calculating the level of confidence in this prediction. The developed sensor combination model showed that a selected combination of sensors could provide more information than any single data source, but that the physical characteristics and use patterns of the monitored space can affect how sensors respond, meaning a generic model to interpret data from multiple spaces was not feasible. The predictive model showed that a trained model could provide a better prediction of near-future occupancy than the typically assumed fixed schedule, up to an average of approximately two hours.
The systems developed in this work were designed to facilitate the proactive control of buildings services, with particular value for slower-response systems such as heating and ventilation. With the application of appropriate control logic, the systems developed can be used to allow for greater energy savings during low or non-occupied periods, while also being more robust to changing occupant patterns and behaviours. |
| first_indexed | 2025-11-14T20:20:48Z |
| format | Thesis (University of Nottingham only) |
| id | nottingham-51466 |
| institution | University of Nottingham Malaysia Campus |
| institution_category | Local University |
| language | English English |
| last_indexed | 2025-11-14T20:20:48Z |
| publishDate | 2018 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | nottingham-514662025-02-28T14:05:37Z https://eprints.nottingham.ac.uk/51466/ Managing the uncertainty of occupant behaviour for building energy evaluation and management Naylor, Sophie The influence of building occupancy and user behaviour on energy usage has been identified as a source of uncertainty in current understanding of operational buildings, and yet it is rarely directly monitored. Gathering data on the occupancy of buildings in use is essential to improve understanding of how energy is used relative to the actual energy requirements of building users. This thesis covers the application of occupancy measurement and processing techniques in order to address the gap in knowledge around the contextual understanding of how occupants’ changing use of a building affects this building’s optimum energy demand in real time. Through targeted studies of running buildings, it was found that typical current occupancy measurement techniques do not provide sufficient context to make energy management decisions. Useable occupancy information must be interpreted from raw data sources to provide benefit: in particular, many slower response systems need information for pre-emptive control to be effective and deliver comfort conditions efficiently, an issue that is highlighted in existing research. Systems utilising novel technologies were developed and tested, targeted at the detection and localisation of occupants’ personal mobile devices, making opportunistic use of the existing hardware carried by most building occupants. It was found that while these systems had the potential for accurate localisation of occupants, this was dependent on personal hardware and physical factors affecting signal strength. Data from these sources was also used alongside environmental data measurements in novel algorithms to combine sensor data into a localised estimation of occupancy rates and to estimate near-future changes in occupancy rate, calculating the level of confidence in this prediction. The developed sensor combination model showed that a selected combination of sensors could provide more information than any single data source, but that the physical characteristics and use patterns of the monitored space can affect how sensors respond, meaning a generic model to interpret data from multiple spaces was not feasible. The predictive model showed that a trained model could provide a better prediction of near-future occupancy than the typically assumed fixed schedule, up to an average of approximately two hours. The systems developed in this work were designed to facilitate the proactive control of buildings services, with particular value for slower-response systems such as heating and ventilation. With the application of appropriate control logic, the systems developed can be used to allow for greater energy savings during low or non-occupied periods, while also being more robust to changing occupant patterns and behaviours. 2018-07-13 Thesis (University of Nottingham only) NonPeerReviewed application/pdf en arr https://eprints.nottingham.ac.uk/51466/7/Corrections_privacy.pdf application/pdf en arr https://eprints.nottingham.ac.uk/51466/1/SN_Thesis_Corrections_fin.pdf Naylor, Sophie (2018) Managing the uncertainty of occupant behaviour for building energy evaluation and management. PhD thesis, University of Nottingham. Building Occupancy Occupancy Measurement Occupancy Prediction Energy Management BEMS Machine Learning Neural Network |
| spellingShingle | Building Occupancy Occupancy Measurement Occupancy Prediction Energy Management BEMS Machine Learning Neural Network Naylor, Sophie Managing the uncertainty of occupant behaviour for building energy evaluation and management |
| title | Managing the uncertainty of occupant behaviour for building energy evaluation and management |
| title_full | Managing the uncertainty of occupant behaviour for building energy evaluation and management |
| title_fullStr | Managing the uncertainty of occupant behaviour for building energy evaluation and management |
| title_full_unstemmed | Managing the uncertainty of occupant behaviour for building energy evaluation and management |
| title_short | Managing the uncertainty of occupant behaviour for building energy evaluation and management |
| title_sort | managing the uncertainty of occupant behaviour for building energy evaluation and management |
| topic | Building Occupancy Occupancy Measurement Occupancy Prediction Energy Management BEMS Machine Learning Neural Network |
| url | https://eprints.nottingham.ac.uk/51466/ |