Abstract—This paper uses MATLAB-Artificial Neural
Networks (ANNs) to predict the current and future energy
consumption of new-builds in the domestic sector. This research
simulates a prototype neighbourhood block, using multiple
prototype domestic sectors across Baghdad, to predict the
future energy use of urban projects and assess the potential use
of renewable sources. This will identify how sustainable
solutions, such as solar energy, may impact on urban
development compared to the conventional methods currently
used. To construct the ANNs, data from one prototype block
from six which were assessed was used. Variables that directly
or indirectly impact on energy consumption were used. The
trained ANN revealed that the use of sustainable solutions, such
as PV systems, can save energy in that there was a 33%
reduction in energy consumption when comparing conventional
and sustainable energy scenarios.
Index Terms—ANN, energy simulated, predict energy, urban
sustainable development.
The authors are with School of Engineering, BRE Centre of Sustainable
Construction, Cardiff University, UK (e-mail: MohsinMM@Cardiff.ac.uk,
beachth@cf.ac.uk, kwan@cardiff.ac.uk).
[PDF]
Cite: Marwah M. Mohsin, Thomas Beach, and Alan Kwan, "Energy Forecasting, Based on ANN Machine Learning,
for Domestic Properties in Dry Hot Arid Regions: A Case
Study in Baghdad," International Journal of Engineering and Technology vol. 10, no. 6, pp. 505-511, 2018.
