Energy consumption forecasting is an operation of predicting the future energy consumption of electricity systems based on past or historical data, which has an increasing impact on society in order to have an accurate forecast and prediction of electricity demand and, for example, to avoid the risk of temporary blackouts or a decrease in power quality. In several countries, these blackouts are a common system failure. This is because these sources (solar power plants, wind power plants) depend on weather conditions, which are stochastic by nature.
Electricity demand forecasting is becoming increasingly important. Correct forecasting makes it possible to plan and expand power sector facilities. Accurate forecasts can save operating and maintenance costs, increase the reliability of the energy supply and delivery system and correct future development decisions.
The increasing share of renewable energy sources in energy production is a rapidly growing field of research, innovation and transfer in recent years.
Surpluses and deficits in energy production negatively affect the operation of electricity grids. For a well-functioning energy system, it is essential to adapt energy production to current demand.
In the daily operation of conventional power plants, adjusting and determining energy production according to demand is a task that can be achieved with accurate forecasting systems. This problem and adjustment, however, becomes more complicated in power plants based on renewable sources. This requires accurate forecasting of energy production from renewable sources, especially when the share of such plants is significant in relation to other sources of electricity. At present, it is necessary to develop methods for forecasting electricity production from wind or solar power plants, or renewable energy in general, taking into account many exogenous factors and, of course, weather conditions.
Existing approaches to forecasting models can be classified into the following four categories: physical, statistical, machine learning and hybrid. Nowadays, thanks to the powerful advancement of intelligent systems, advanced Deep Learning models have become an indispensable tool in the realization of accurate forecasting models.
Topics. Suggested topics for papers include, but are not limited to, the following topics:
- - Prediction in renewable energy
- - Energy forecasting
- - Machine learning in time series prediction
- - Time series analysis with computational intelligence
- - Integration of system dynamics and forecasting models in energy problems
- - Advanced Methods and On-Line Learning In Time Series
- - Hierarchical forecasting
- - Deep learning in Time Series Forecasting for Energy problem
- - New advances in Long short-term memory network
- - Transformer-based model for forecasting time series data
Organizers:
Prof. Peter Glösekötter, Peter Glösekötter is Professor at the Department of Electrical Engineering and Computer Science at FH Münster and researcher in the field of electronic and information technology, contributing extensively to advancements in sensors, magnetometers, and energy storage systems. His work, published across a variety of prestigious journals and conference proceedings, showcases a strong focus on nanotechnology, quantum sensing, and machine learning applications for robust measurement and diagnosis tools. Innovative approaches are evidenced by his leadership in developing zinc-air batteries and pioneering energy-autonomous systems, significantly impacting smart health and environmental resource management.
Dr. Joseph Moerschell, Joseph Moerschell is Professor of Electronics and Mechatronics in HES-SO, the University of Applied Sciences of Western Switzerland, in Sion. After a PhD in industrial electronics at EPFL, Lausanne, he specialized in instrumentation development for science applications, both in space missions and in extreme environment applications in alpine and polar regions. This includes micro-mechanical manipulators, innovative sensors, optic and electromagnetic, and associated signal processing and autonomous power supply. More recently, his group develops laser spectroscopy and quantum sensing to probe such environments.
Prof. Ignacio Rojas, Ignacio Rojas is Professor at the Department of Computer Engineering, Automation and Robotics at the University of Granada. His field of research regards the study of complex multidimensional systems using intelligent systems, supported by high-performance computing platforms, focused on solving real problems in various fields, such as bioinformatics, biomedicine, and time series prediction, among others. As a result of the research developed, he has published more than 270 contributions reflected in Clarivate Web of Science, has participated with papers in more than 125 international conferences related to his field of research, has directed 27 doctoral theses, has organized a total of 23 international and 5 national conferences.
Prof. Tilman Sanders, Tilman Sanders is Professor at the Department of Electrical Engineering and Computer Science at FH Münster and researcher in the field of power electronics and power systems engineering. His work focuses on increased energy efficiency by utilizing new power semiconductor materials like SiC and GaN in power converters for drives and renewables and on the integration of renewable power sources as well as e-mobility and heat pumps into the power grid, while maintaining its stability.
Prof. Markus Gregor, Markus Gregor is Professor at the Department of Engineering Physics at the FH Münster University of Applied Sciences. One particular focus of his quantum technology group is the implementation of quantum sensors based on NV centers in diamond. The main areas in this research are cost-efficient and brilliant single photon sources based on NV centers in nanodiamonds, functionalization of optical fibers with micro- and nanodiamonds for nanodiamonds for sensor technology, microstructuring of optical waveguides with embedded nanodiamonds, Recently, his research also included the use of machine learning for quantum sensor applications.
Prof. Sarah Trinschek, Sarah Trinschek is Professor at the Münster University of Applied Sciences. Her research activities involve modelling and simulation of physical systems, advanced data analysis and machine learning. With a background in nonlinear physics, complex systems and soft matter, her recent research interests focus on projects in collaboration with experimental partners. These involve the development of intelligent sensor systems by the application of machine learning to sensor data as well as the control and optimization of optical systems.
Prof. Ruxandra Stoean, Ruxandra Stoean is Associate Professor at the Department of Computer Science, Faculty of Sciences, University of Craiova, Romania. She holds a PhD in Computer Science, with a thesis on optimization through evolutionary computation. Her current research interests involve the development of deep learning models for images and signals, with applications in engineering and renewable energy. She is doctoral co-supervisor at the Inter-University PhD Program “Electric Energy Systems” at the University of Malaga, Spain.