CURRENT APPROACHES AND NEW RESEARCH IN
MODERN SCIENCES
International scientific-online conference
20
MODERN APPROACHES TO ASSESSING THE POTENTIAL OF
PHOTOVOLTAIC SYSTEMS ON BUILDING ROOFTOPS USING GIS,
LIDAR, AND MACHINE LEARNING
Rakhmatov Odiljon Adkhamjon ugli
Fergana Polytechnic Institute, Republic of Uzbekistan
https://doi.org/10.5281/zenodo.15023707
Keywords:
GIS, LiDAR, photovoltaic systems, solar potential, machine
learning, optimization, modeling.
Abstract:
In the context of the global transition to renewable energy
sources (RES), solar energy plays a key role in the energy strategies of many
countries, including Uzbekistan. One promising direction is the installation of
photovoltaic (PV) systems on building rooftops, which allows for the efficient
use of urban areas for electricity generation. However, assessing the solar
potential of rooftops is a complex task that requires considering multiple factors,
such as roof geometry, shading, orientation, and tilt angle. This article provides
an overview of modern methods for assessing the solar potential of rooftops
using geographic information systems (GIS), LiDAR data, and machine learning.
Key research findings:
1.
GIS and LiDAR technologies:
GIS combined with LiDAR data enables
accurate reconstruction of 3D models of urban areas, taking into account
shading, tilt angle, and roof orientation. This makes them key tools for assessing
the technical potential of solar energy in cities. For example, studies conducted
in Piedmont (Italy) and Seoul (South Korea) found that the use of solar panels on
rooftops could cover up to 30% of the cities' annual energy consumption11.
2.
Machine learning:
Machine learning methods, such as Support Vector
Machine (SVM), Random Forests, and deep neural networks, significantly
improve the accuracy of assessments and automate the analysis process. For
instance, in a study conducted in Switzerland, the average error of estimates was
only 9.5%, while in China, the application of deep learning achieved 94%
accuracy in detecting the presence of solar panels on rooftops22.
3.
Optimization of PV system placement:
Optimization algorithms, such as
Integer Linear Programming (ILP) and Proximal Policy Optimization (PPO),
provide economically justified distribution of PV systems. They consider budget
constraints, electricity costs, panel degradation, and changes in pricing factors,
thereby improving project profitability. For example, in a study conducted in
Hong Kong, the use of ILP optimization improved the performance-cost ratio by
17.7% compared to the best Monte Carlo alternatives33.
CURRENT APPROACHES AND NEW RESEARCH IN
MODERN SCIENCES
International scientific-online conference
21
4.
Shading analysis:
Hillshade analysis and 3D modeling play a crucial role
in assessing solar potential. For example, in studies conducted in South Korea,
the actual solar energy generation was only 12% of the physical potential due to
building shading44.
5.
Practical application:
The practical application of these methods includes
the development of national strategies for integrating solar energy, urban
planning, and energy system optimization. The results can be used by
governments to develop subsidy programs for PV systems and to accurately
forecast solar energy generation in specific areas55.
Recommendations for Uzbekistan:
1.
Application of GIS and LiDAR data:
To assess the solar potential of
rooftops in Uzbekistan, it is recommended to use GIS and LiDAR data, which
allow for precise determination of roof tilt, orientation, and shading. This is
particularly important for cities with high building density, such as Fergana.
2.
Use of machine learning methods:
Machine learning methods,
such as SVM and Random Forests, can be used for automatic roof classification
and assessment of their suitability for PV system installation. This will
significantly improve the accuracy of calculations and automate the analysis
process.
3.
Optimization of PV system placement:
For optimal distribution of
solar panels on building rooftops, it is recommended to use optimization
algorithms, such as ILP and genetic algorithms. This will allow for consideration
of budget and network constraints, increasing project profitability.
4.
Shading analysis:
For accurate assessment of solar potential, it is
recommended to use Hillshade analysis and 3D modeling, which take into
account the impact of shading on electricity generation.
5.
Development of urban energy strategy:
Based on the conducted
research, it is recommended to develop an urban energy strategy that includes
assessing the impact of mass PV system installation on the energy system load
and developing scenarios for integrating solar generation into the power grid.
Conclusion:
The review of scientific works on assessing the potential of photovoltaic
systems on building rooftops shows that the application of GIS, LiDAR, satellite
data, machine learning, and optimization methods significantly improves the
accuracy of solar potential assessment. The use of these technologies allows for
consideration of spatial, technical, and economic factors affecting the efficiency
of PV systems, as well as automating the design process. For Uzbekistan,
CURRENT APPROACHES AND NEW RESEARCH IN
MODERN SCIENCES
International scientific-online conference
22
especially for cities with high building density such as Fergana, these methods
can form the basis for developing strategies to integrate solar energy into urban
infrastructure.
References:
1.
Abdo Abdullah Ahmed Gassar, Seung Hyun Cha, Review of geographic
information systems-based rooftop solar photovoltaic potential estimation
approaches at urban scales, Applied Energy, Volume 291, 2021, 116817, ISSN
0306-2619.
2.
Seunghoon Jung, Jaewon Jeoung, Hyuna Kang, Taehoon Hong, Optimal
planning of a rooftop PV system using GIS-based reinforcement learning,
Applied Energy, Volume 298, 2021, 117239, ISSN 0306-2619.
3.
Haoshan Ren, Zhenjun Ma, Antoni B. Chan, Yongjun Sun, Optimal planning
of municipal-scale distributed rooftop photovoltaic systems with maximized
solar energy generation under constraints in high-density cities, Energy, Volume
263, Part A, 2023, 125686, ISSN 0360-5442.
4.
Elham Fakhraian, Marc Alier Forment, Francesc Valls Dalmau, Alireza
Nameni, Maria José Casañ Guerrero, Determination of the urban rooftop
photovoltaic potential: A state of the art, Energy Reports, Volume 7, Supplement
3, 2021, Pages 176-185, ISSN 2352-4847.
5.
Mesude Bayrakci Boz, Kirby Calvert, Jeffrey R. S. Brownson. An automated
model for rooftop PV systems assessment in ArcGIS using LIDAR, AIMS Energy,
2015, 3(3): 401-420.
