Multi-Agent Negotiation Algorithms for Resources Cost Estimation: A Case Study

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 1997

Autores

Fonseca, JM (Autor) Outra; Eugénio Oliveira (Autor) FEUP; Garção, AS (Autor) Outra

Ata de Conferência Internacional

Título: Progress in Artificial Intelligence, 8th Portuguese Conference on Artificial Intelligence, EPIA '97, Coimbra, Portugal, October 6-9, 1997, Proceedings

Páginas: 195-206

Indexação

Scopus - 0 Citações

Outras Informações

ID Authenticus: P-008-NW7

DOI: 10.1007/BFb0023922

Abstract (EN): In this paper we describe appropriate negotiation protocols we have developed for a multi-agent system whose aim is to advice on distributed resources selection, as well as cost estimation, in big civil construction companies. Prices of each specific resource have an influence on the results of the negotiation algorithm which leads to the best agents (resources) coalition for a given task. Moreover, several interesting problems have been addressed in order to make agents negotiation strategies more competitive and adequate. This is the case of the ¿self depreciation problem¿ which arises when an agent (a resource) belongs to more than one coalition which alters its own behaviour during the competition between those coalitions. Our negotiation protocol and agents strategies are able to provide the company with a precise idea about the respective importance of each resource and reach a solution, i.e. find the best set of resources, to deal competitively with the task in hands. Another important factor our approach takes into account during agents negotiation, is the influence of time (date of task announcement, starting and ending task dates) which may guide each agent negotiation strategy according with a forecast of their own employment possibilities in the future. We conclude the paper with a description of a realistic example in the civil construction domain, which illustrates the concepts we have developed. © Springer-Verlag Berlin Heidelberg 1997.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Nº de páginas: 12