Application of Genetic Algorithms to Distributed Optimization Problems under Fuzzy Constraints

In this work, we are interested in a variant of DisCSPs which is distributed optimization problems under fuzzy constraints (DisFCSPs: Distributed Fuzzy Constraint Satisfaction Problems). This problem lies at the confluence of two areas of research: Distribute d optimization and fuzzy optimization. F...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Procedia computer science Jg. 159; S. 1258 - 1266
Hauptverfasser: Medini, Ghofrane, Bouamama, Sadok
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 2019
Schlagworte:
combinatorial optimization
Distributed Fuzzy Constraint Satisfaction Probl ems
Genetic Algorithm
metaheuristics
Multi-Agent Systems
combinatorial optimization
metaheuristics
Multi-Agent Systems
Genetic Algorithm
Distributed Fuzzy Constraint Satisfaction Probl ems
ISSN:1877-0509, 1877-0509
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In this work, we are interested in a variant of DisCSPs which is distributed optimization problems under fuzzy constraints (DisFCSPs: Distributed Fuzzy Constraint Satisfaction Problems). This problem lies at the confluence of two areas of research: Distribute d optimization and fuzzy optimization. Fuzzy optimization by metaheuristics, and more particularly by genetic algorithms, is well studied in the centralized case, contrary to the decentralized case. We present a new multi-agent framework designed to create fuzzy distributed constraint satisfaction (DisFCSP) problems and their resolution using population based metaheuristics. Different types of agents (interface agents, mediating agent, proposing agents and negotiating agents) interact in a cooperative architecture to receiv e, read and deliver information about problems, coordinate activities, build promising solutions and generate improved solutions, realizing a metaheuristic process as a collaborative behavior. The genetic algorithm (GA) is able to exploit the architecture and can be easily described in the proposed framework.
ISSN:1877-0509
1877-0509
DOI:10.1016/j.procs.2019.09.295