Machine Learning for Sustainable Power Systems
ECML 2024 Workshop
Vilnius, on the 9th September 2024
2:00 PM - 6:00PM
Room Tau
Embedded Files
Submission Closed & Notification Send
Workshop Objective
Workshop Objective
Introduction
The Machine Learning for Sustainable Power Systems (ML4SPS) workshop aims to bring together researchers in the fields of machine learning and energy systems at the European Conference of Machine Learning (ECML PKDD).
Energy systems around the world are facing the challenge of integrating massive amounts of new renewable energy production. Unlike traditional energy production methods, new renewables are highly variable, decentralized and only partially predictable. At the same time, new consumption patterns and additional uses of electricity are emerging, such as self-consumption, heat pumps or electric vehicles. These new consumers increase the complexity of the grids in terms of monitoring and control, but also allow for new flexibility in the future. The use of machine learning methods can help to exploit these new flexibilities to overcome the increasing complexity of modern energy systems. To this end, machine learning can take advantage of technological advances such as smart meters and novel communication devices.
Consequently, ML4SPS aims to bring together researchers and practitioners to share their knowledge and experience in the field of renewable energy systems, grid management and machine learning. The workshop itself will be organised by experts in the field of renewable energy systems and machine learning who will share their knowledge and experience with the participants. The workshop will be interactive and will include the presentation of the submitted papers, a poster session and two keynote speakers from the field of machine learning in energy systems.
Research Topics
Research Topics
The workshop ML4SPS focuses on, but is not limited to, several key topics within the energy system, as listed below. However, contributions to other related research areas are also welcome. In terms of methods and algorithmic approaches, the workshop does not limit any contributions as long as they are from the field of machine learning.
Forecasting in energy systems
Advanced predictive models for forecasting of electricity demand, wind energy, and solar/PV power generation using deep learning and time series analysis.
Ensemble methods and uncertainty quantification in energy forecasting to enhance decision-making processes.
Grid Control and Optimization:
Machine learning algorithms for real-time grid management and operation, including load balancing, outage management, frequency regulation, grid calculations and voltage control.
Integration of distributed energy resources (DERs) and energy storage solutions through intelligent control systems.
Approaches on generalizability und adaptability of power system applications
Robustness and Trust in Industrial AI Applications:
Methods to enhance the robustness of machine learning models against adversarial attacks, data corruption, and model drift in energy systems.
Strategies for building trust and ensuring transparency in AI-based decision-making, including explainable AI (XAI) techniques for energy applications.
Asset Management and Predictive Maintenance:
Predictive analytics for asset health monitoring, leveraging IoT data, forecasting equipment failures.
Machine learning frameworks for life-cycle management of energy infrastructure, enhancing reliability and optimizing operational costs.
Techniques for real-time anomaly detection and diagnosis in power systems.
Energy Markets and Trading Strategies:
Application of machine learning in analysing energy market dynamics, price forecasting, and risk assessment
AI-driven tools for optimizing energy procurement, portfolio management, and demand-side response.
Impact analysis of renewable energy penetration on energy markets.
Energy Integration and Planning:
Optimization models for the effective integration of renewable energy source
Simulation and scenario analysis AI-tools with respect to grid planning, renewable energy deployment and infrastructure resilience.
Integration of Electric Vehicles into the Power Grid:
Machine learning for managing the impact of electric vehicle charging on the power grid.
Smart charging algorithms and strategies for electric vehicles.
Ethical AI and Socioeconomic Impacts of Machine Learning in Power Systems:
Ethical considerations and governance frameworks for the deployment of AI in energy systems
Machine learning's role in addressing energy poverty and ensuring equitable access to clean energy.
Workshop Program
Workshop Program
Call for Papers
Call for Papers
Introduction and workshop objectives
Integrating large amounts of renewable energy into the energy system is a major challenge. Unlike traditional energy production methods, renewable energy (wind and solar) is highly variable, decentralized and only partially predictable. In addition, energy consumption is changing rapidly with the increasing use of heat pumps and electric cars. This variability can lead to imbalances between energy supply and demand, which can strain the grid and increase the risk of blackouts.
Machine Learning (ML) has emerged as a promising tool for addressing the challenges of integrating renewable energy into energy systems. ML algorithms can be used in various fields of to improve the forecasting accuracy of renewable energy production, optimize energy planning and network control, enable predictive maintenance of energy infrastructure, design efficient energy markets or assess the variability of the energy consumption. However, in the future use of possible ML algorithms in energy systems, it is crucial to consider robustness and to ensure confidence in their results, as the models will be used in critical infrastructures. Therefore, trust for industrial ML methods needs to be considered to ensure reliable and effective models.
The Machine Learning for Sustainable Power Systems workshop aims to bring together researchers and practitioners in the fields of machine learning and energy systems to discuss the latest advancements in machine learning for sustainable power systems applications. The workshop will provide a platform for researchers to present their work, exchange ideas, and network with other experts in the field.
Target Audience
The workshop welcomes submissions from researchers, practitioners, and PhD students working in the areas of machine learning, energy systems, and sustainable energy. Papers are encouraged from a variety of perspectives, including:
Forecasting in energy systems, e.g., demand, wind and solar forecasting
Grid control, grid analysis and optimization
Robustness and trust for industrial AI applications
Asset management and predictive maintenance
Energy Markets and trading strategies
Energy integration and planning
Integration of electric vehicles into the power grid
Ethical AI and socioeconomic impacts of ML in power systems
Submission Guidelines
Authors are invited to submit original papers (maximum 16 pages for technical papers, maximum 8 pages for short position papers) describing their research on machine learning applications for sustainable power systems. In both cases the page limit including references, for which there is no limit. Overlength papers will be rejected without review (papers with smaller page margins and font sizes than specified in the author instructions and set in the style files will also be treated as overlength). Papers should be submitted electronically in PDF format by 15 June 2024 and shall prepared and submitted according to LNCS Format. Authors are encouraged to include open-source code with their work.
Authors who submit their work to ML4SPS 2024 commit themselves to present their paper at the workshop in case of acceptance (at least one author). ML4SPS 2024 considers the author list submitted with the paper as final. No additions or deletions to this list may be made after paper submission, either during the review period, or in case of acceptance, at the final camera ready stage. At least one author of each accepted paper must have a full registration and be in Vilnius to present the paper.
Papers without a full registration or in-presence presentation won’t be included in the post-workshop Springer proceedings.
Paper Evaluation
Submitted papers will be reviewed by a panel of experts in the fields of machine learning and energy systems. The review process will be single-blind. Papers will be evaluated based on their innovativeness, technical accuracy, and potential to advance the field. Successful submissions will be presented at the workshop and published on the ML4SPS website. Additionally, a Best Paper Award will be given to the paper that makes the most significant contribution to machine learning for sustainable power systems.
Post-workshop proceedings
As stipulated by ECML 2024 organizers, "the Workshops and Tutorials will be included in a joint Post-Workshop proceeding published by Springer Communications in Computer and Information Science, in 1-2 volumes, organised by focused scope and possibly indexed by WOS. Papers authors will have the faculty to opt-in or opt-out".
Important Dates:
Paper submission deadline: 15 June 2024 Extended to 30 June 2024
Notification of acceptance: 15 July 2024 Extended to 30 July 2024
Camera-Ready version due: 15 August 2024 Extended to 15 Sept 2024
Submit via Microsoft CMT:
https://cmt3.research.microsoft.com/ECMLPKDDWorkshops2024/Track/19/Submission/Create
Please be sure to select ML4SPS from the drop down menu when creating your submission. Authors who submit their work to ML4SPS 2024 commit themselves to present their paper at the workshop in Vilnius in case of acceptance (at least one author).
Keynote Speakers
Keynote Speakers
Damien Ernst
Damien Ernst received the M.Sc. and Ph.D. degrees in engineering from the University of Liège, Belgium, in 1998 and 2003, respectively. He is currently Full Professor at the University of Liège and Visiting Professor at Télécom Paris. His research interests include electrical energy systems and reinforcement learning, a subfield of artificial intelligence. He is also the CSO of Haulogy, a company developing intelligent software solutions for the energy sector. He has co-authored more than 300 research papers and two books. He has also won numerous awards for his research and, among which, the prestigious 2018 Blondel Medal. He is also regularly consulted by industries, governments, international agencies and the media for its deep understanding of the energy transition.
For more information visit blogs.ulg.ac.be/damien-ernst/
Keynote: The four-layer decision-making problem for power system operation: How can AI help?
Electrical power systems are evolving rapidly nowadays to integrate renewables and new electrical loads, such as heat pumps or electric vehicles. However, many experts believe their evolution is still too slow, and that they represent a major bottleneck for rapidly decarbonizing our societies. One reason behind this is that power system engineers grapple with the complexity of these systems, which are increasingly structured around four layers: microgrids, distribution networks, transmission networks, and supergrids. Each layer presents its own complex decision-making problems and needs to be coordinated with the other ones through various mechanisms to ensure a reliable and affordable power supply. In this talk, Professor Damien Ernst will describe some of the decision-making challenges faced when operating these layers and the multiple opportunities they offer for AI researchers seeking challenging problems that can make a difference in the energy transition.
Keynote material: https://damien-ernst.be/2024/09/06/the-four-layer-decision-making-problem-forpower-system-operation-how-can-ai-help/
Pedro P. Vergara
Pedro P. Vergara is an assistant professor at the Intelligent Electrical Power Grids (IEPG) group at the Delft University of Technology, in the Netherlands. He is also a Senior Member of the IEEE from the Power and Energy Society. From 2019 to 2020, he was with the Eindhoven University of Technology, in the Netherlands as a postdoctoral researcher, in which he was involved in several Dutch-funded projects with industry and academic partners. He holds a Ph.D. degree from the University of Campinas, in Brazil, and from the University of Southern Denmark, in Denmark. Dr. Pedro P. Vergara devotes his research to developing new mathematical programming models and data-driven control approaches to operate electrical distribution systems with high penetration of low carbon energy resources (i.e. PVs, EVs, electric heat-pumps, etc). Currently, he is leading developments in the H2020 MAGPIE project, NWO DATALESS project and NWO ALIG4Energy project, which involve several Dutch and European industry and academic partners. He has more than 30 scientific publications in international leading conferences and journals. He also serve as Guest Editor in special issues related to active distribution networks in the IEEE MPCE Journal and IEEE Transactions on JPV Journal.
For more information visit www.pedropvergara.nl/
Keynote: Advances in Reinforcement Learning for Distribution Systems Operation ― Enforcing Safety and Ensuring Scalability
(Deep) Reinforcement Learning (RL) presents a promising avenue for optimizing distribution systems operation. RL’s strength lies in its ability to learn distribution systems’ stochastic behavior via extensive interactions, usually using advanced simulators. Although promising, several challenges remain unresolved to enable the wide adoption of RL-based decisions in the energy sector. Two of these challenges relate to the capabilities of Rl algorithms in enforcing operational constraints in real-time as well as ensuring action space scalability. In this talk, we will dive into recent advances in mathematical programming and deep learning theory to ensure safety in RL deployment as well as the use of advanced state representation to ensure RL scalability.
Workshop Chairs
Workshop Chairs
Organization Committee
Malte Lehna
Fraunhofer Institute for Energy Economics and Energy System Technology IEE
Clara Holzhüter
Intelligent Embedded Systems, University Kassel
Clément Goubet
Data Science and AI Team, RTE, France's Transmission System Operator
Dr. Christoph Scholz
Fraunhofer Institute for Energy Economics and Energy System Technology IEE
Intelligent Embedded Systems, University Kassel
Prof. Dr. Bernhard Sick
Intelligent Embedded Systems, University Kassel
Program Committee
The international program committee for the workshop Machine Learning for Sustainable Power Systems comprises prominent experts from the research field of energy systems as well as state-of-the-art machine learning domains. This ensures a thorough and equitable reviewing process.
Ada Diaconescu: Autonomous and Critical Embedded Systems, Telecom Paris, France
Alexander Dreher: Energy Economics & Energy System Technologies, Fraunhofer IEE, Germany
Alexander Scheidler: Grid Operation Management, Fraunhofer IEE, Germany
Alice Moallemy-Oureh: Graphs in AI and Neural Networks, University of Kassel, Germany
Alessandro Zocca: Department of Mathematics, Vrije Universiteit Amsterdam, Netherlands
Benjamin Donnot: Data Science and AI tech lead, RTE, France
Benjamin Säfken: Institute of Mathematics, TU Clausthal, Germany
Björn Hoppmann: Cognitive Energy Systems, Fraunhofer IEE, Germany
Catalina Obando: Data Science and AI team member, RTE, France
Chandana Priya Nivarthi: Autonomous Energy Systems, University of Kassel, Germany
Jan Viebahn: Data Scientist, TenneT TSO BV, Netherlands
Jean Thorey: Data Science and AI team member, RTE, France
Karim Chaouache: Data Science and AI team member, RTE, France
Laura Hansel, Neural Data Science Group, University of Göttingen, Germany
Manuel Wickert: Energy Informatics, Fraunhofer IEE, Germany
Martin Braun: Energy Management and Power System Operation, University of Kassel, Germany
Mohamed Hassouna: Autonomous Energy Systems, University of Kassel, Germany
Nikos Bosse: Forecasting Expert, London School of Hygiene & Tropical Medicine, Great Britain
Paul Almasan: Telefónica Research, Spain
Pawel Lytaev: Energy Management and Power System Operation, University of Kassel, Germany
Pedro Vergara Barrios: Intelligent Electrical Power Grids, TU Delft, Netherlands
René Heinrich: RL for Cognitive Energy Systems, University of Kassel, Germany
Sven Tomforde: Intelligent Systems, Christian-Albrechts-University Kiel, Germany
Thomas Kneib: Chair of Statistics, Georg August-Universität Göttingen, Germany
Viktor Eriksson Möllerstedt: Data Scientist, Svenska kraftnät, Sweden
Additional Committee Members to be announced
Sponsors
Sponsors
The workshop Machine Learning for Sustainable Power Systems is sponsored by RTE, France's Transmission System Operator. We thank the sponsor for making this workshop possible.
Page updated
Google Sites
Report abuse