Affiliations: Dipartimento di Elettrotecnica, (DEI) Politecnico di Milano, Milano, Italy | Department of Computer Science and Engineering, University of Minnesota, 4-192 EE/CS Building, 200 Union Street S.E., Minneapolis, MN 55455-0159, USA
Correspondence:
[*]
Corresponding author: Dr. N. Gatti, E-mail: [email protected].
Abstract: The deregulation of the electricity markets produced significant economic benefits. Improving their efficiency is a prominent scientific challenge. We focus on wholesale electricity markets, in which generators sell electricity to a public agency by means of a central auction. The multi-agent literature studies these markets according to two main approaches, each one providing a different level of expressiveness. The first approach, based on game theory, provides a formal analysis of the markets, allowing one to find the optimal generators' strategies in simple market models (e.g., omitting the auction mechanism). The second approach, based on multi-agent simulations, assumes that generators implement simple learning algorithms. This approach allows one to tackle complex market models, but no formal result on the optimality of the solution is provided in the literature. In this paper, we provide an algorithmic game theory study that improves the state of the art related to both the two previous approaches. Concerning the first approach: we enrich the game theoretic models available in the literature by introducing the auction mechanism, we provide an algorithm to solve the auction winner determination problem and an efficient algorithm to compute generators' optimal strategies. Concerning the second approach: we formally study the dynamics of learning generators analyzing the convergence to the optimal strategies.
