This paper reports on results from an agent-based simulation
model that comprises three interrelated markets in the electricity sector:
a day-ahead electricity market, a market for balancing power, and a carbon
exchange for CO2 emission allowances. Agents seek to optimize trading
strategies over the two electricity markets through reinforcement learning;
they also integrate market results from emissions trading into their
reasoning. Simulation outcomes show that the model is able to closely
reproduce observed prices at the German power markets for the analysis
period of 2006. The model is thus applicable for analyzing different market
designs in order to derive evidence for policy advice; one example for such
an analysis is given in this contribution.
Dieser Eintrag ist Teil der Universitätsbibliographie.
Search Authors in