Agent-Based Modeling of Wholesale Electricity Market

The article investigates the free electricity market via agent-based modeling. The aim is to create a simple and quite plausible simulation model of the market, where suppliers and buyers, participating in the bilateral auction, learn to submit the most profitable bids. Whereas this kind of simulation models has been developed for the EU and USA markets by foreign researchers, for the Russian market they have not been made yet. The suggested theoretical agent-based model of interaction in the day-ahead market uses Erev and Roth learning algorithm and allows us to calculate and analyze equilibrium price, volume, social welfare and its distribution between buyers and sellers of electricity. The modeling demonstrates that it is possible to lower prices and redistribute social welfare in favor of buyers, provided that the learning agents can submit any bids.