Information Technology Reference
In-Depth Information
economic agents with learning capabilities. In order to overcome limitations in the sign
of payoff typical of reinforcement learning algorithms proposed in the literature, an en-
hanced version of the Roth-Erev algorithm (i.e., which takes into account positive, null
and negative payoffs) has been presented and discussed. Furthermore, due to its com-
plex high-voltage transmission network, the Italian power exchange (IPEX) has been
taken as case of study. This resulted in replicating the exact market clearing procedure
and considering generation plants in direct correspondence with the real ones. Results
on the convergence of the enhanced Roth-Erev learning algorithm pointed out effective-
ness of the proposed algorithm. In particular, the evolution of the strategy probabilities
pointed out different groups of agents characterized by different convergence rates that
strongly depend on the role of the agent in the market. This fact confirms the direct ap-
plicability of the proposed Enhanced Roth-Erev learining algorithm for economic and
financial applications. Moreover, computational experiments of the ABM IPEX model
performed within the GAPEX pointed out a close agreement with historical data during
both peak- and off-peak load hours, confirming the direct applicability of the GAPEX
to model and to simulate power exchanges in particular for what-if analysis and market
design.
Acknowledgements.
E. Guerci and M.A. Rastegar collaborated to the design and the
development of the GAPEX framework. This work has been partially supported by the
University of Genoa, by the Italian Ministry of Education, University and Research
(MUR) under grant PRIN 2007, by the European Social Fund (ESF) and by Regione
Liguria, Italy.
References
1. Bagnall, A., Smith, G.: A multi-agent model of the UK market in electricity generation. IEEE
Transactions on Evolutionary Computation 9(5), 522-536 (2005)
2. Ball, P.: The earth simulator. New Scientist 2784, 48-51 (2010)
3. Bower, J., Bunn, D.W.: Experimental analysis of the efficiency of uniform-price versus dis-
criminatory auctions in the England and wales electricity market. Journal of Economic Dy-
namics & Control 25, 561-592 (2001)
4. Bunn, D.W., Oliveira, F.: Agent-based simulation: an application to the new electricity
trading arrangements of England and wales. IEEE Transactions on Evolutionary Compu-
tation 5(5), 493-503 (2001)
5. Camerer, C., Ho, T.: Experience-weighted attraction learning in normal-form games. Econo-
metrica 67, 827-874 (1999)
6. Cau, T.D.H., Anderson, E.J.: A co-evolutionary approach to modelling the behaviour of par-
ticipants in competitive electricity markets. In: IEEE Power Engineering Society Summer
Meeting, vol. 3, pp. 1534-1540 (2002)
7. Cincotti, S., Guerci, E., Raberto, M.: Agent-based simulation of power exchange with het-
erogeneous production companies. Computing in Economics and Finance 2005, Society for
Computational Economics 334 (2005)
8. GME: Official web site (2010),
9. GME: Uppo auction module user manual, appendix a - market splitting auction algorithm.
Tech. rep., GME (2010),
http://www.mercatoelettrico.org/En/
