Abstract
With growing trends towards smaller-scale, low carbon energy systems such as microgrids and later generation district heating, coupled with the increased use of building simulation in bottom-up stock modelling, there is increasing interest in being able to automatically generate multiple, high-resolution profiles of both occupancy and occupancy-dependent household demand. This has resulted in the emergence of a range of tools capable of producing time-varying occupancy and demand profiles that capture many of the characteristics evident in real-world data. However, as a result of limited data availability, these tools have typically been calibrated using composite data from multiple individuals or households. This results in the production of profiles which whilst statistically faithful to the average characteristics of the underpinning data, fail to capture the variance seen from household-to-household in the real world. This paper attempts to address this shortcoming, developing and testing a new method within an existing model structure for the production of occupancy and linked demand data using a probabilistic model that is representative of the overall population from which it is derived, but which has an improved ability to generate specific outputs that better match the behaviour of specific households. The new approach utilises previously established occupancy and demand modelling methods, along with the composite population-based calibration data used within the models. However, the temporal predictive basis of the models has been manipulated to account for individual behaviours whilst retaining the overall statistical characteristics of the source data. This was achieved by adapting the Markov chain timing basis for previously developed models and factoring the probability values. A linked electrical demand model has also been adapted by manipulation of the relative timing of use of different appliances and demands to account for differences between individual and average behaviours. The described approach has the benefit of not requiring any additional calibration data, which for both occupancy and energy demand is often scarce. The predictions of the improved model and previous version are compared to real occupancy and demand data, indicating that the alterations enable significantly more diverse profiles to be generated, whilst still being representative of the supporting data.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.