Existing Power Network as a Challenge
The insufficient accuracy of current forecasting models for volatile electricity production, such as wind and solar energy, leads to significant discrepancies between predicted and actual loads. These discrepancies not only strain grid stability but also cause additional costs for end users.
Using AI Models to Predict More Efficient Utilization
POWERCAST aims to bridge this gap with adaptive AI models. The project focuses on two main applications: ensuring compliance with the N-1 safety criterion in the power grid through more accurate load forecasts using adaptive AI models, and detecting extreme situations in the grid.
The project will develop novel AI models based on adaptive learning that integrate various information sources. By collaborating with leading players in the energy industry and accessing large energy data sources, these methods will be benchmarked and validated.
Lower Energy Costs and Leverage New Energy Sources
The POWERCAST project has the potential to generate financial benefits and improve the sustainability of the energy system. By developing more precise forecasting models, system-wide costs could potentially be reduced and the integration of renewable energies facilitated. The project is clearly aligned with the EU’s goals to significantly increase the share of renewable energy and achieve climate neutrality by 2050, and will help to address current and future challenges in power grid optimization.
Consortium: Concentrated Expertise from Austria
The consortium consists of Linz Netz GmbH, Austrian Power Grid AG, HAKOM Time Series GmbH, the Energy Institute at Johannes Kepler University Linz, and RISC Software GmbH. It is funded by the Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology, represented by the Austrian Research Promotion Agency (FFG), through the AI for Green 2023 call. The consortium combines expertise and prior experience in various fields such as data management, data analysis, and energy economics and management.
The energy-related use cases and associated data sources are provided by Linz Netz GmbH and Austrian Power Grid AG. HAKOM Time Series GmbH is responsible for setting up a data warehouse where data from different sources will be merged, validated, and harmonized. RISC Software GmbH is in charge of project coordination and the development of the forecasting models based on machine learning, including the necessary data preprocessing. The Energy Institute at Johannes Kepler University Linz complements the consortium with its expertise in developing and evaluating the POWERCAST model and considering the legal framework.
More Information
Images
© RISC Software GmbH, reproduction free of charge
