Please use this identifier to cite or link to this item:
|Type:||Artigo de evento|
|Title:||Electricity Market Simulation: Multiagent System Approach|
|Abstract:||This paper suggests a multiagent system (MAS) approach for market simulation. This is achieved through analysis, modeling, implementation and simulation of artificial markets populated by software agents that represent economic self interested agents. Software agents are the constructs of a complex system, an artificial market that model a real existing market or an outline of a market design. The interest in simulating a market is multiple: exploiting existing market rules, searching for market design flaws and loopholes, and supporting decision making during a market mechanism design process. The main aim of the suggested approach is to analyze the behavior that emerges from the interaction of self interested agents acting in an artificial market. AEMAS (Artificial Economy MultiAgent System), a multiagent system architecture inspired by the Market Oriented Programming (MOP) approach is defined. In different economical sectors, e.g. energy markets, there is no consensus about which structures lead to social welfare maximization outcomes. An approach to find adequate architectures allows different market structure instances to be created and simulated, to ease the design and analysis of alternative structures. These alternatives can then be compared and potential design flaws eventually risen by simulation identified. Taking the electricity market as an example, two instances of the proposed architecture are presented, corresponding to the centralized dispatch arrangement common to non restructured markets, and the auction based pool, common to restructured markets. Copyright 2008 ACM.|
|Appears in Collections:||Unicamp - Artigos e Outros Documentos|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.