Stochastic Optimization Problems in Trading with Renewable Energies

Power generation in Germany and worldwide has been shifting more and more towards renewable energies for some years now.

This green turnaround has an impact on the energy markets: the volatility associated with some renewable energy producers such as wind farms or solar panels is also reflected in prices.

It is therefore very important for distributors of, for example, wind farm electricity to be as accurate as possible in their prediction of future feed-in.


For price sensitive buyers, on the other hand, such a forecast can be an indicator of when prices will be lower or higher than usual. If they have the opportunity to flexibly cover their needs within a certain time frame, they will of course try to do so as cost-effectively as possible.

It becomes exciting when a consumer is also a producer ("prosumer") and has the possibility to generate electricity himself within a certain framework. Of course, the prosumer can still buy electricity on the market. Now the question is how the prosumer should behave in the best possible way in order to both reliably cover his demand and pay the lowest possible price for the electricity.

To analyze this situation, both classical methods of stochastic optimization (model-based) and newer methods (data-based) are used. A game theoretical approach can also be assumed. The goals are optimal trading recommendations for consumers or the development of price strategies for the supplier, ideally leading to an equilibrium model for the market.