Advancing AI Integration in Smart Distribution Grids
About
TEF DSO is a future-ready testing environment that aims to support Distribution System Operators (DSOs) integrate AI-driven solutions into their existing infrastructures. This includes interoperability with modern automation architectures and orchestration layers, that in turn allow real-world testing for service providers aiming to tailor their products to the needs of DSOs.
Testing Infrastructure
The node is based in Germany and comprises four partners. RWTH provides the ACS real-time simulation laboratory, while GridHound and EsGeht! are responsible for offering smart solutions and AI services for energy management and grid operation. SWW is a DSO offering communication and transportation services and will act as an end user.
Physical
The physical infrastructure is provided by RWTH, including the framework that connects the resources with the virtual assets.
Germany
- A Real Time Digital Simulator (RTDS) hosted by the ACS laboratory equipped with:
- GTIO, GTNET, GTSYNC and GTFPGA extension cards
- 4 OPAL-RT Technologies devices
- OPAL-RTs eMEGASIM, ePHASORSIm, eFPGASIM and HYPERSIM simulation tools
- VILLAS framework that connects computational and HiL resources, including:
- A wind turbine testing facility
- A grid-connected medium voltage dc grid
- External and internal HPC facilities
Virtual
The virtual assets are also provided by RWTH and include various frameworks for managing and handling data from the resources.
- VILLAS real-time co-simulation infrastructure for local and external resources
- SOGNO microservice platform for AI service orchestration
- Digital Twins of power distribution systems
- An energy data space instance based on ENERSHARE
Available Data
This node provides dynamic and static synthetic real-time data related to the distributions systems of the DSO, including:
- Instantaneous electric variables
- Synchronized AC and DC measurements of distribution systems
- EV charging data
- Power and energy flows
- Load profiles and renewable generation profiles