Prediction and realization of the energy provision with decentralized power plants


The project PREmdeK demonstrates with realistic data how the green factory is possible for specific companies.



A decentralized power plant is a summary of local, renewable energy-producers, such as wind power, photovoltaic, or solar heat. These should cover the entire power- and/or heat-demand of the (manufacturing) company.

Self-supplied and CO2-neutral

The production facility itself and the decentralized power plant together form a local energy system. How this local; autarkic and CO2-neutral supply is possible; what elements are additionally important; and how everything works together are demonstrated with real case studies (refer also to our publications).

The main result of the project is the development of the PREmdeK Simulation. The PREmdeK Simulation calculates the possible energy generation by the renewable power plant based on weather forecast data (historic or forecast) and supplies the energy demand of the consumer for the 15 min time interval and for a specific timeframe.

Future development of the simulation “PREmdeK 3.0”

The main change made to the software is the substitution of the monolithic structure by a multi-agent structure. The single agents of the hybrid power plant in PREmdeK 2.0 are transformed into a multi-agent type in PREmdeK 3.0.

Any agent sharing an ontology with the rest of the participating agents is able to take part in the simulation. This makes the software scalable as well as future-proof, because new components can be added easily and older ones can be modified effortlessly.

Other changes introduced in the software are the distribution of energy according to sustainable criteria (e.g. energy costs, CO2 emission, corporate image) and the addition of new agent types for energy production.

PREmdeK 3.0: In progress since 2016

PREmdeK 2.0: Completed (2011 - 2014)

PREmdeK 1.0: Completed (2011 - 2014)


Student projects and final theses

Aitor Casado Garcia (Industrial Informatics): Functional specifications for PREMDEK 3.0: A multi-agent simulation software for the analysis of virtual power plants (Masterarbeit 2016)

Isabella Anna Wera (Medieninformatik): Visualisierung eines quasi autarken Energiesystems mit der Simulationssoftware AnyLogic (Praxisprojekt 2016)

Nino Feldermann (IBS): Vergleich zweier Simulationsmodelle anhand der Erstellung wirtschaftlicher alternativer Energiesysteme für ein Industrieunternehmen (Bachelorarbeit 2018)



Dr.-Ing. Agnes Pechmann

Research assistants

Maximilian Zarte (Department of Mechanical Engineering)

Stefan Woltmann (Department of Computer Science)

Student assistants

Sunita Gill (Bioinformatic)


The further development of the simulation PREmdeK is currently being tested and evaluated. Student and industrial projects are used to verify the results in different scenarios and to optimize the simulation. Concrete results of the projects can be found in recent publications. For questions regarding the simulation or ongoing interest in a project, please contact us.