Replicator agents for electricity markets

Author

Ono, Teruo ; Verghese, George C.

Author_Institution

Lab. for Electromagn. & Electron. Syst., Massachusetts Inst. of Technol.

fYear

2005

fDate

6-10 Nov. 2005

Abstract

The main focus of this paper is to explore the dynamic behavior of an auction system for an electricity market. In order to cope with this complex problem, agent-based simulation has been previously used, where autonomous agents learn through the results of repeated auctions. In this paper, the replicator equation is introduced as a learning algorithm that can be applicable to agent-based simulation. A repeated electricity auction based on a Japanese power market is simulated with different strategy selection algorithms incorporating mechanisms for learning. The results are compared to show the applicability of the replicator equation

Keywords

learning (artificial intelligence); multi-agent systems; power markets; power system economics; Nash equilibrium; agent-based simulation; auction system; autonomous agents; electricity auction; electricity markets; learning algorithm; repeated auctions; replicator agents; replicator equation; Analytical models; Autonomous agents; Dynamic programming; Electricity supply industry; Equations; Laboratories; Power markets; Power system dynamics; Power system modeling; Power system simulation;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Systems Application to Power Systems, 2005. Proceedings of the 13th International Conference on

Conference_Location

Arlington, VA

Print_ISBN

1-59975-174-7

Type

conf

DOI

10.1109/ISAP.2005.1599310

Filename

1599310