Please use this identifier to cite or link to this item: http://hdl.handle.net/1959.13/921037

OpenURL Link
Title
New robust multiobjective ant colony optimization (MOACO) method
Author/Creator
Mortazavi Naeini, S. Mohammad;  Kuczera, George;  Cui, Lijie
Description
Most decision problems require tradeoffs between conflicting objectives. The aim of multiobjective optimization is to define the Pareto optimal set of solutions that is to be used in the trade-off process. Recently several heuristic methods have been developed to find the Pareto set; these include the ε-dominance multiobjective evolutionary algorithm (εMOEA) and the multiobjective ant colony optimization (MOACO) algorithm. In this study a new MOACO method is proposed to overcome the shortcomings of earlier MOACO algorithms. This method is more robust, uses fewer tuning parameters, and is independent of the number of objectives and independent of the problem. The key improvements involve only updating pheromone on non-dominated solutions, using a time decay factor to avoid excessive pheromone build-up and switching off heuristic information. A case study based on stormwater harvesting for potable reuse illustrates the superiority of the new MOACO method. Benchmarking the new MOACO against εMOEA revealed similar performance. Both produced similar diversity in Pareto sets. εMOEA tended to dominate MOACO solutions for function evaluations up to 10,000 with MOACO then becoming dominant.
Relation
9th International Conference on Hydroinformatics (HIC 2010). Proceedings of the 9th International Conference on Hydroinformatics 2010 (Tianjin, China 7 - 10 September, 2010) p. 2285-2292
Relation
http://www.iwahq.org/n1
Date
2010
Publisher
Chemical Industry Press (CIP)
Keyword(s)
MOACO;  multiobjective ant colony optimization;  ε-dominance multiobjective evolutionary algorithm;  decision making
Resource Type
conference paper
Identifier
http://hdl.handle.net/1959.13/921037
Identifier
ISBN:9787894723246
Language
eng
Reviewed
Reviewed
84 Visitors 97 Hits 0 Downloads
Save/E-mail Citation
Citation Format
E-mail Address
Subject
"New robust multiobjective ant colony optimization (MOACO) method"
 
OR