Project

Back to overview

Hybrid renewable energy potential for the built environment using big data: forecasting and uncertainty estimation

English title Hybrid renewable energy potential for the built environment using big data: forecasting and uncertainty estimation
Applicant Scartezzini Jean-Louis
Number 167285
Funding scheme NRP 75 Big Data
Research institution Laboratoire d'énergie solaire et physique du bâtiment EPFL - ENAC - IIC - LESO-PB
Institution of higher education EPF Lausanne - EPFL
Main discipline Civil Engineering
Start/End 01.05.2017 - 31.07.2021
Approved amount 586'228.00
Show all

All Disciplines (4)

Discipline
Civil Engineering
Other disciplines of Engineering Sciences
Architecture and Social urban science
Theoretical Physics

Keywords (7)

Built environment; Hybrid renewable energy; Forecasting and uncertainty; Sustainability; Machine learning; Intelligent decision-making; LiDAR point cloud

Lay Summary (German)

Lead
Es ist nicht leicht, das Potenzial eines städtischen Gebiets für hybride erneuerbare Energien (Hybrid Renewable Energy Potential: HyREP) einzuschätzen. Ziel dieses Projekts ist die Entwicklung eines neuartigen Konzepts für HyREP-Prognosen in der Schweiz, das sich auf Big-Data-Technologien und moderne maschinelle Lernverfahren stützt.
Lay summary

Gebäude haben den grössten Anteil am Energiebedarf der Schweiz. Sie verbrauchen mehr als 40 % der gesamten Energie und mehr als 32 % der Elektrizität. Eine Senkung des Energieverbrauchs und der Treibhausgasemissionen setzt voraus, dass unsere Gebäude bedeutend energieeffizienter werden und vorrangig erneuerbare Energien nutzen. Die Einschätzung des Potenzials für einen kombinierten Einsatz erneuerbarer Energien in Gebäuden wird daher für die Schweiz immer wichtiger – vor allem für Gemeinden, Hauseigentümer und öffentliche Energieversorger. Solche Einschätzungen liefern nützliche Informationen über das Potenzial, das in der Energieerzeugung sowie bei Sparmassnahmen vorhanden ist.

HyREP-Systeme kombinieren Solar- und Windenergie mit oberflächennaher geothermischer Energie. Solche Systeme tragen in Gebäuden dazu bei, die Grösse von autonomen Energiesystemen, die notwendige Energiespeicherkapazität und die Gesamtbetriebskosten beträchtlich zu reduzieren. Mit Blick auf die Energiewende ist es wichtig, das Potenzial für den kombinierten Einsatz von erneuerbaren Energien in Gebäuden genauer zu erfassen und die Entscheide von Anspruchsgruppen sowie die Bauvorschriften in der Schweiz durch datengestützte Verfahren zu untermauern.

Dank der im Rahmen dieses Projekts eingesetzten Big-Data-Technologien können wir erstmals das Energiesparpotenzial für Gebäude in der gesamten Schweiz einschätzen. Die Ergebnisse werden sehr wahrscheinlich die Energiepolitik der städtischen Regionen in der Schweiz beeinflussen. Die Methoden können aber auch für andere Länder von Nutzen sein. Mithilfe der Datenbank für erneuerbare Energien in Gebäuden (Building Renewable Energy Database) lassen sich die Energieeinsparungen und die Energieversorgung einzelner Gebäude oder ganzer Siedlungen überall in der Schweiz visualisieren.

Direct link to Lay Summary Last update: 26.07.2017

Lay Summary (French)

Lead
Estimer le potentiel des systèmes hybrides d’énergies renouvelables (HyREP) dans les zones urbaines est difficile. L’objectif du projet est de développer une nouvelle approche cohérente pour évaluer l’HyREP en Suisse au moyen des technologies issues des Big Data et des méthodes modernes d’apprentissage automatique.
Lay summary

Les bâtiments représentent la plus grande part de la demande en énergie en Suisse. Ils constituent plus de 40% de la demande globale et plus de 32% de la demande en électricité. Afin de réduire la consommation d’énergie et les émissions de gaz à effet de serre, ils doivent devenir plus efficients et miser en priorité sur les sources d’énergie renouvelable. Estimer l’HyREP pour l’environnement construit est en conséquence devenu un enjeu extrêmement important en Suisse, particulièrement pour les municipalités, les propriétaires immobiliers et les services publics. De telles estimations fournissent des informations très utiles sur la production d’énergie et les économies d’énergie potentielles.

Les systèmes HyREP combinent diverses sources d’énergies renouvelables : solaire, éolienne et géothermique de surface. Dans le secteur du bâtiment, de telles installations réduisent de façon substantielle la taille des systèmes énergétiques autonomes, la capacité nécessaire de stockage de l’énergie et l’ensemble des coûts d’exploitation. Dans la perspective du virage énergétique, il est important d’évaluer le potentiel combiné des ressources en énergie pour l’environnement construit en utilisant des mégadonnées. Cela aidera les acteurs concernés du bâtiment en Suisse à définir leurs stratégies et prendre leurs décisions.

En utilisant les technologies Big Data, notre projet permettra d’évaluer pour la première fois dans quelle mesure la demande en énergie des bâtiments peut être réduite en Suisse. Les résultats sont susceptibles d’influencer les politiques énergétiques urbaines dans notre pays. Les méthodes fondées sur les données pourront toutefois aussi être utilisées dans d’autres pays. La base de données sur l’énergie renouvelable dans le bâtiment nous permettra de visualiser les économies et l’approvisionnement en matière d’énergie pour des bâtiments individuels ou des groupes de bâtiments partout en Suisse.

Direct link to Lay Summary Last update: 26.07.2017

Lay Summary (English)

Lead
Estimating the hybrid renewable energy potential (HyREP) of an urban area is challenging. The aim of this project is to develop a novel, coherent approach to forecasting HyREP in Switzerland using Big Data technologies and contemporary machine learning methods.
Lay summary

Buildings represent the largest share of the energy demand in Switzerland. They account for more than 40% of the overall energy demand and more than 32% of the electricity demand. To reduce energy consumption and greenhouse gas emissions, buildings need to become much more energy-efficient and to rely primarily on renewable energy resources. Accordingly, estimating HyREP for the built environment is becoming a very important issue in Switzerland, particularly for municipalities, building owners and public utilities. Such estimates provide very useful information regarding potential energy generation and energy savings.

HyREP systems combine solar, wind and shallow geothermal energy. In the building sector, such systems substantially reduce the size of standalone energy systems, the required energy storage capacity and total operating costs. In view of the energy transition, it is important to assess the combined potential of renewable energy resources for the built environment using Big Data to support stakeholders’ decisions and policies for buildings in Switzerland.

Using Big Data technologies, our project will make it possible to assess for the first time how much of the energy demand from buildings can be reduced throughout Switzerland. The results are likely to influence urban energy policies in this country, but the data-driven methods can also be used for other countries. The Building Renewable Energy Database will allow us to visualise energy savings and supply for individual buildings as well as for groups of buildings anywhere in Switzerland.


Direct link to Lay Summary Last update: 26.07.2017

Responsible applicant and co-applicants

Employees

Project partner

Publications

Publication
Uncertainty quantification in extreme learning machine: Analytical developments, variance estimates and confidence intervals
Guignard Fabian, Amato Federico, Kanevski Mikhail (2021), Uncertainty quantification in extreme learning machine: Analytical developments, variance estimates and confidence intervals, in Neurocomputing, 456, 436-449.
Quantifying the technical geothermal potential from shallow borehole heat exchangers at regional scale
Walch Alina, Mohajeri Nahid, Gudmundsson Agust, Scartezzini Jean-Louis (2021), Quantifying the technical geothermal potential from shallow borehole heat exchangers at regional scale, in Renewable Energy, 165, 369-380.
A novel framework for spatio-temporal prediction of environmental data using deep learning
Amato Federico, Guignard Fabian, Robert Sylvain, Kanevski Mikhail (2020), A novel framework for spatio-temporal prediction of environmental data using deep learning, in Scientific Reports, 10(1), 22243-22243.
Spatio-temporal evolution of global surface temperature distributions
Amato Federico, Guignard Fabian, Humphrey Vincent, Kanevski Mikhail (2020), Spatio-temporal evolution of global surface temperature distributions, in CI2020: 10th International Conference on Climate Informatics, virtual United KingdomNIL, NIL.
Advanced Analysis of Temporal Data Using Fisher-Shannon Information: Theoretical Development and Application in Geosciences
Guignard Fabian, Laib Mohamed, Amato Federico, Kanevski Mikhail (2020), Advanced Analysis of Temporal Data Using Fisher-Shannon Information: Theoretical Development and Application in Geosciences, in Frontiers in Earth Science, 8(255), 1-11.
Advanced Analysis of Temporal Data Using Fisher-Shannon Information: Theoretical Development and Application in Geosciences
GuignardFabian, LaibMohamed, AmatoFederico, KanevskiMikhail (2020), Advanced Analysis of Temporal Data Using Fisher-Shannon Information: Theoretical Development and Application in Geosciences, in Frontiers in Eatrh Science, 8(255), 1-8.
Analysis of air pollution time series using complexity-invariant distance and information measures
Amato Federico, Laib Mohamed, Guignard Fabian, Kanevski Mikhail (2020), Analysis of air pollution time series using complexity-invariant distance and information measures, in Physica A: Statistical Mechanics and its Applications, 547, 124391-124391.
Analysis of air pollution time series using complexity-invariant distance and information measures
Amato Federico, Laib Mohamed, Guignard Fabian, Kanevski Mikhail (2020), Analysis of air pollution time series using complexity-invariant distance and information measures, in Physica A: Statistical Mechanics and its Applications, 547, 124391-124391.
Big data mining for the estimation of hourly rooftop photovoltaic potential and its uncertainty
WalchAlina, CastelloRoberto, MohajeriNahid, ScartezziniJean-Louis (2020), Big data mining for the estimation of hourly rooftop photovoltaic potential and its uncertainty, in Applied Energy, 262, 1-18.
Analysis of temporal properties of extremes of wind measurements from 132 stations over Switzerland
Telesca Luciano, Guignard Fabian, Laib Mohamed, Kanesvki Mikhail (2020), Analysis of temporal properties of extremes of wind measurements from 132 stations over Switzerland, in Renewable Energy , 145, 1091-1103.
A critical comparison of methods to estimate solar rooftop photovoltaic potential in Switzerland
Walch Alina, Mohajeri Nahid, Scartezzini Jean-Louis (2019), A critical comparison of methods to estimate solar rooftop photovoltaic potential in Switzerland, in Journal of Physics: Conference Series, 1343, 012035-012035.
Wind profile prediction in an urban canyon: a machine learning approach
MaureeDasaraden, CastelloRoberto, ManciniGianluca, NuttaTullio, ZhangTianchu, ScartezziniJean-Louis (2019), Wind profile prediction in an urban canyon: a machine learning approach, in Journal of Physics: Conference Series, 1343(012047), 1-6.
Wavelet Scale Variance Analysis of Wind Extremes in Mountainous Terrains
TelescaLuciano, GuignardFabian, HelbigNora, KanevskiMikhail (2019), Wavelet Scale Variance Analysis of Wind Extremes in Mountainous Terrains, in Energies, 12(16), 1-10.
A machine learning approach for mapping the very shallow theoretical geothermal potential
AssoulineDan, MohajeriNahid, GudmundssonAgust, ScartezziniJean-Louis (2019), A machine learning approach for mapping the very shallow theoretical geothermal potential, in Geothernal Energy, 7(19), 1-50.
Wavelet variance scale-dependence as a dynamic discriminating tool in high-frequency urban wind speed time series
Guignard Fabian, Mauree Dasaraden, Kanevski Mikhail, Telesca Luciano (2019), Wavelet variance scale-dependence as a dynamic discriminating tool in high-frequency urban wind speed time series, in Physica A, 525, 771-777.
Investigating the time dynamics of wind speed in complex terrains by using the Fisher–Shannon method
Guignard Fabian, Lovallo Michele, Laib Mohamed, Golay Jean, Kanevski Mikhail, Helbig Nora, Telesca Luciano (2019), Investigating the time dynamics of wind speed in complex terrains by using the Fisher–Shannon method, in Physica A: Statistical Mechanics and its Applications, 523, 611-621.
Integrating urban form and distributed energy systems: Assessment of sustainable development scenarios for a Swiss village to 2050
Mohajeri Nahid, Perera Dasun, Coccolo Silvia, Mosca Lucas, Le Guen Morgane, Scartezzini Jean-Louis (2019), Integrating urban form and distributed energy systems: Assessment of sustainable development scenarios for a Swiss village to 2050, in Renewable Energy, 143, 810-826.
A solar-based sustainable urban design: The effects of city-scale street-canyon geometry on solar access in Geneva, Switzerland
Mohajeri N., Gudmundsson A., Kunckler T., Upadhyay G., Assouline D., Kämpf J.H, Scartezzini J.L. (2019), A solar-based sustainable urban design: The effects of city-scale street-canyon geometry on solar access in Geneva, Switzerland, in Applied Energy, 240, 173-190.
Community detection analysis in wind speed-monitoring systems using mutual information-based complex network
Laib Mohamed, Guignard Fabian, Kanevski Mikhail, Telesca Luciano (2019), Community detection analysis in wind speed-monitoring systems using mutual information-based complex network, in Chaos, 29(043107), 1-12.
Linearity versus non-linearity in high frequency multilevel wind time series measured in urban areas
Mauree Dasaraden, Kanevski Mikhail, Telesca Luciano, Laib Mohamed, Guignard Fabian (2019), Linearity versus non-linearity in high frequency multilevel wind time series measured in urban areas, in Chaos, Solitons & Fractals, 120, 234-244.
Fisher–Shannon Complexity Analysis of High-Frequency Urban Wind Speed Time Series
Guignard Fabian, Mauree Dasaraden, Lovallo Michele, Kanevski Mikhail, Telesca Luciano (2019), Fisher–Shannon Complexity Analysis of High-Frequency Urban Wind Speed Time Series, in Entropy, 21(1), 47-47.
A Fast Machine Learning Model for Large-Scale Estimation of Annual Solar Irradiation on Rooftops
Walch Alina, Castello Roberto, Mohajeri Nahid, Scartezzini Jean-Louis (2019), A Fast Machine Learning Model for Large-Scale Estimation of Annual Solar Irradiation on Rooftops, in ISES Solar World Congress 2019/IEA SHC International Conference on Solar Heating and Cooling for Bui, Santiago, ChileInternational Solar Energy Society, Freiburg i.B., Germany.
Deep learning in the built environment: automatic detection of rooftop solar panels using Convolutional Neural Networks
CastelloRoberto, RoquetteSimon, EsguerraMartin, GuerraAdrian, ScartezziniJean-Louis (2019), Deep learning in the built environment: automatic detection of rooftop solar panels using Convolutional Neural Networks, in Journal of Physics: Conference Series, 1343(012034), 1-6.
Machine learning and geographic information systems for large-scale wind energy potential estimation in rural areas
AssoulineDan, MohajeriNahid, Dasaraden Mauree, ScartezziniJean-Louis (2019), Machine learning and geographic information systems for large-scale wind energy potential estimation in rural areas, in Journal of Physics: Conference Series, 1343(012036), 1-6.
Spatio-temporal modelling and uncertainty estimation of hourly global solar irradiance using Extreme Learning Machines
WalchAlina, CastelloRoberto, MohajeriNahid, GuignardFabian, KanevskiMikhail, ScartezziniJean-Louis (2019), Spatio-temporal modelling and uncertainty estimation of hourly global solar irradiance using Extreme Learning Machines, in Energy Procedia, 158, 6378-6383.
A city-scale roof shape classification using machine learning for solar energy applications,
Mohajeri Nahid, AssoulineDan, GuiboudBernard, BillA., GudmundssonA., ScartezziniJean-Louis (2018), A city-scale roof shape classification using machine learning for solar energy applications,, in Renewable Energy, 81-93.
Combining Fourier Analysis and Machine Learning to Estimate the Shallow-Ground Thermal Diffusivity in Switzerland
AssoulineDan, MohajeriNahid, GudmundssonAgust, ScartezziniJean-Louis (2018), Combining Fourier Analysis and Machine Learning to Estimate the Shallow-Ground Thermal Diffusivity in Switzerland, in IGARSS, (8517938), 1144-1147.
Improving the energy sustainability of a Swiss village through building renovation and renewable energy integration
Le Guen Morgane, MoscaLucas, PereraDasun, CoccoloSilvia, MohajeriNahid, ScartezziniJean-Louis (2018), Improving the energy sustainability of a Swiss village through building renovation and renewable energy integration, in Energy and Buildings, 906-923.

Collaboration

Group / person Country
Types of collaboration
Institute of Geography and Sustainability, UNIL Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Research Infrastructure
Lund University Sweden (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Amstein + Walthert Geneva SA Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Industry/business/other use-inspired collaboration
Institute of Methodologies for Environmental Analysis (IMAA) Italy (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Exchange of personnel
Centre for Advanced Modelling Science (CADMOS) Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Research Infrastructure
Measurement of Turbulence in an Urban Setup (MoTUS) Switzerland (Europe)
- Research Infrastructure
Sustainable Urban Development Programme, University of Oxford Great Britain and Northern Ireland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Exchange of personnel
- Industry/business/other use-inspired collaboration
SCCER FEEB&D P-II WP3 Energy Performance at Regional and National Scale’ Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Research Infrastructure
- Industry/business/other use-inspired collaboration
University of Basel Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
Learn to Forecast (L2F) Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Industry/business/other use-inspired collaboration
Swiss Federal Institute for Forest, Snow and Landscape Research (WSL) Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
EPFL Scientific IT and Application Support (SCITAS) Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Research Infrastructure

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
28th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning Talk given at a conference On feature selection using anisotropic general regression neural network 02.10.2020 Bruges, Belgium Kanevski Mikhail; Guignard Fabian;
10th International Conference on Climate Informatics Talk given at a conference Spatio-temporal evolution of global surface temperature distributions 23.09.2020 NIL, Great Britain and Northern Ireland Guignard Fabian; Kanevski Mikhail;
ISES Solar World Congress 2019 Talk given at a conference A fast machine learning model for large-scale estimation of annual solar irradiation on rooftops 04.11.2019 Santiago, Chile Walch Alina;
AMLD 2019 Applied Machine Learning Days Talk given at a conference Machine Learning and Environmental Risk 26.01.2019 Lausanne, Switzerland Kanevski Mikhail;
SGM 2018 16th Swiss Geoscience Meeting Talk given at a conference Mutual information-based complex network for wind speed in Switzerland 30.11.2018 Berne, Switzerland Kanevski Mikhail; Guignard Fabian;
SGM 2018 16th Swiss Geoscience Meeting Talk given at a conference Application of the Fisher-Shannon plane to high frequency wind speed in Switzerland 30.11.2018 Berne, Switzerland Kanevski Mikhail; Guignard Fabian;
IAMG 2018 International Association for Mathematical Geosciences Talk given at a conference Clustering of environmental data using local fractality concept and machine learning 02.09.2018 Olomouc, Czech Republic Guignard Fabian; Kanevski Mikhail;
ICAE 2019 International Conference on Applied Energy Talk given at a conference Spatio-temporal modelling and uncertainty estimation of hourly global solar irradiance using Extreme Learning Machines 22.08.2018 Hongkong, Hongkong Walch Alina;
GeoENV 2018 Geostatistics for Environmental applications Talk given at a conference Quantification of Extreme Learning Machine modelling uncertainty using bootstrapping 03.07.2018 Belfast, Ireland Kanevski Mikhail; Guignard Fabian;
EGU 2018 European Geosciences Union General Assembly 2018 Talk given at a conference A spatio-temporal model to estimate hourly solar radiation using Extreme Learning Machines 09.04.2018 Vienne, Austria Walch Alina;
JBGE 2018 Journées Biennales des Geosciences et de l'Environnement Poster Estimation de l’HyREP en zones urbaines à l’aide du Machine Learning 12.02.2018 UNIL Campus, Lausanne, Switzerland Kanevski Mikhail; Guignard Fabian; Walch Alina;
SGM 2017 15th Swiss Geoscience Meeting, Davos, Talk given at a conference Uncertainty quantification in environmental data driven modeling using machine learning 17.11.2017 Davos, Switzerland Guignard Fabian; Kanevski Mikhail;
CISBAT 2017 International Conference – Future Building and Districts : Energy Efficiency from Nano to Urban Scale Talk given at a conference Extending building integrated photovoltaïcs (BiPV) using distributed energy-hubs : a case study in Cartigny/GE Switzerland 06.09.2017 EPFL Campus, Lausanne, Switzerland Scartezzini Jean-Louis; Mohajeri Pour Rayeni Nahid;


Self-organised

Title Date Place
AMLD 2019 Applied Machine Learning Days 26.01.2019 Lausanne, Switzerland

Knowledge transfer events

Active participation

Title Type of contribution Date Place Persons involved
Applied machine learning days 2019 / Track AI & Environment Performances, exhibitions (e.g. for education institutions) 26.01.2019 Lausanne, Switzerland Castello Roberto; Walch Alina;


Self-organised

Title Date Place

Communication with the public

Communication Title Media Place Year
Media relations: print media, online media Energies: 50% des toits suisses peuvent produire de l'électricité Le Nouvelliste Western Switzerland 2020
Talks/events/exhibitions Event Note: Match-Making in Big Data with Academia and Industry German-speaking Switzerland Italian-speaking Switzerland Rhaeto-Romanic Switzerland Western Switzerland 2020
Media relations: print media, online media Im Winter fehlt der Strom – wie gross ist das Potenzial des Solarstroms? NZZ am Sonntage German-speaking Switzerland 2020
Media relations: print media, online media Jedes zweite Dach in der Schweiz wäre für Solaranlagen geeignet NZZ German-speaking Switzerland 2020
Media relations: radio, television La moitié des toits suisses peuvent produire du courant Radio Lac Western Switzerland 2020
Media relations: radio, television La moitié des toits suisses pourraient produire de l'électricité RTS Western Switzerland 2020
Media relations: print media, online media L'énergie photovoltaïque a du potentiel en Suisse 24Heures Western Switzerland 2020
Media relations: print media, online media Les données éclairent les toits suisses bulletin.ch Western Switzerland 2020
New media (web, blogs, podcasts, news feeds etc.) Neue Studie zeigt: Jedes zweite Dach in der Schweiz taugt für Solarzellen watson International 2020
New media (web, blogs, podcasts, news feeds etc.) Study measures Switzerland's potential geothermal heating capacity EPFL News International 2020
Media relations: print media, online media What if half of Switzerland's rooftops produced electricity? EPFL News International 2020
New media (web, blogs, podcasts, news feeds etc.) Can Big Data revolutionize the Energy System? NRP75 BigData Dialog Platform International Western Switzerland German-speaking Switzerland 2019
New media (web, blogs, podcasts, news feeds etc.) CISBAT2019 – Energy Efficiency and Renewables meet Big Data NRP 75 Bigdata Dialog International 2019
New media (web, blogs, podcasts, news feeds etc.) The HyEnergy project: Swiss renewable energy potential in the digital era NRP75 BigData Dialog Platform Western Switzerland Italian-speaking Switzerland International 2019
New media (web, blogs, podcasts, news feeds etc.) Track “AI & Environment” at AMLD 2019 NRP75 BigData Dialog Platform Western Switzerland German-speaking Switzerland International 2019

Associated projects

Number Title Start Funding scheme
167285 Hybrid renewable energy potential for the built environment using big data: forecasting and uncertainty estimation 01.05.2017 NRP 75 Big Data

Abstract

Buildings have the largest share in energy demand in Switzerland. Building stocks in Switzerland use more than 40% of the overall energy demand in Switzerland and more than 32% of the electricity demand in the country. In order to reduce the energy demand and greenhouse gas emissions, buildings need to become much more energy efficient and the energy demand should be primarily (increasingly) satisfied through renewable energy resources (e.g. solar energy, wind micro-turbine, and geothermal heat pumps). Forecasting renewable energy potential for the built environment, therefore, is becoming an important issue, particularly for municipalities, building owners, and energy suppliers so as to obtain meaningful information regarding potential energy generation and potential energy saving. However, the stochastic nature of the energy resources (particularly solar and wind), massive amounts of heterogeneous, semi-structured and unstructured data generated at unprecedented scale, as well as the complexity of disparate data sources make it extremely challenging to forecast spatio-temporal potential of renewable energy resources. This includes high dimensional complex environmental data as well as massive cloud of LiDAR point data for measuring an entire built environment (buildings, trees, and landscape). Data-driven approaches and machine learning algorithms will play an important role to overcome these challenges. Given this context, the project aims at (i) estimating hybrid renewable energy potential in the built environment, rather than focusing on stand-alone energy potential, in order to mitigate the effects of variability in the individual energy resources and improve the reliability of the power generation, (ii) developing machine learning algorithms for spatio-temporal environmental data processing and analysis as well as for LiDAR point cloud urban dataset classification, (iii) applying the developed algorithms based on Extreme Learning Machine (ELM), to the built environment for predicting energy generation and potential energy savings of hybrid renewable resources, (iv) analysing the forecasting models according to the projected climate scenarios for 2035 and 2050, (v) uncertainty estimations and model validation using measurement data from weather stations and energy provider companies (e.g. EPFL campus, Geneva, Zurich, Basel), (vi) proposing Building Renewable Energy Database (BRED), geo-visualisation tools and renewable energy mapping to support evidence-based decision-making processes.
-