.png){kind=icon}
.png){kind=icon}
Return
Using agents for solving a multi-commodity-flow problem
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ejor.2008.01.002.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We investigate a commodity trading problem in a flow network with arbitrary topology where sinks combine commodities into bundles in order to generate profits. Our focus is the profit maximization problem for the trading network under both central and distributed control. We compute solutions for the central control problem using an integer linear program while we compute solutions for the distributed case by implementing the nodes in the network as software-agents that exchange messages in order to establish profitable trades. We report on computational results using both methods and demonstrate that there is a connection between agent profits and a centrality measure developed for the problem. We also demonstrate that with our current agent strategy, there is a trade-off between the agents acting too quickly before enough information is available and waiting too long and thus giving each agent too much information and thus too much power over the outcome. (C) 2008 Elsevier B.V. All rights reserved.
Keywords:
Distributed decision making
Multi-agent systems
Integer programming
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
6
Papers:
2.2W
Citations:
6.4W