A smart urban energy prediction system to support energy planning in the residential sector

The UK residential sector accounts for approximately 27%., and 17% of the country energy consumption and its CO2 emission, respectively. Thus, developing appropriate polices to reduce the environmental factors, which are associated with the CO2 emissions of a rapidly growing urban population, constitutes a high priority. Moreover, ensures the creation of cities that respect the natural environment and the well-being of future generations. While a great deal of expertise on detailing and constructing low-energy buildings and cities has been developed, it is fragmented and does not consider the concept of household life-cycle demographic transitions in the prediction of residential energy consumption. This research aimed to develop an integrated 3D urban energy prediction tool which supports decision-making for a sustainable energy monitoring and planning in the residential sector. This, while considering the CK household demographic transition patterns in the energy prediction process. To attain the above aim, the research embraced a mixed-methods methodology with 4 stages of practical implementations. In stages 1 and 2, statistical procedures such as binary logistic regression, were applied to the British household panel data survey (BHPS) to attain the two following objectives. First, to analyse the socio-economic and demographic factors affecting the UK household transitions; consequently, predict future transition patterns in the next 10-15 years. Secondly, to investigate the impact of the predicted transition patterns on the residential energy consumption. The examination of the findings indicated that the nature of independent factors and their degree of influence on household transition patterns were not consistent across the 10-15 years. Moreover, it advised that household transitions mostly have a positive but weak effect on their energy usage. Based on those findings, a linear regression model was developed to predict the households' future electricity usage in function of their transition, demographic and socio­ economic variables. In phases 3 and 4, a 3D urban energy prediction tool (EvoEnergy) was developed by first building a 3D semantic model of a pilot area in Nottingham city. Moreover, by integrating the research findings from stages 1 and 2 into EvoEnergy using computer scripting, open-source game technology, and 3D visualisation techniques. Finally, despite the facts that the benchmarking of EvoEnergy highlighted some areas for improvement, it has advised that EvoEnergy has the ability to predict domestic electricity consumption at the building and neighbourhood levels with a good accuracy(+/- 5% error)