A Unified Evolutionary Optimization Procedure for Single, Multiple, and Many Objectives

Traditionally, evolutionary algorithms (EAs) have been systematically developed to solve mono-, multi-, and many-objective optimization problems, in this order. Despite some efforts in unifying different types of mono-objective evolutionary and non-EAs, researchers are not interested enough in unify...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on evolutionary computation Jg. 20; H. 3; S. 358 - 369
Hauptverfasser: Seada, Haitham, Deb, Kalyanmoy
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York IEEE 01.06.2016
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
Algorithm design and analysis
Algorithms
Design engineering
Design optimization
Equivalence
Evolutionary
Evolutionary algorithms
Heuristic algorithms
mono-objective optimization
NSGA-III
Objectives
Optimization
Optimization algorithms
Representations
Sociology
Software
Software algorithms
Statistics
Unified algorithms
mono-objective optimization
multiobjective optimization
non-dominated sorting genetic algorithm (NSGA)-III
Many-objective optimization
unified algorithms
ISSN:1089-778X, 1941-0026
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Traditionally, evolutionary algorithms (EAs) have been systematically developed to solve mono-, multi-, and many-objective optimization problems, in this order. Despite some efforts in unifying different types of mono-objective evolutionary and non-EAs, researchers are not interested enough in unifying all three types of optimization problems together. Such a unified algorithm will allow users to work with a single software enabling one-time implementation of solution representation, operators, objectives, and constraints formulations across several objective dimensions. For the first time, we propose a unified evolutionary optimization algorithm for solving all three classes of problems specified above, based on the recently proposed elitist, guided nondominated sorting procedure, developed for solving many-objectives problems. Using a new niching-based selection procedure, our proposed unified algorithm automatically degenerates to an efficient equivalent population-based algorithm for each class. No extra parameters are needed. Extensive simulations are performed on unconstrained and constrained test problems having single-, two-, multi-, and many-objectives and on two engineering optimization design problems. Performance of the unified approach is compared to suitable population-based counterparts at each dimensional level. Results amply demonstrate the merit of our proposed unified approach and motivate similar studies for a richer understanding of the development of optimization algorithms.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:1089-778X
1941-0026
DOI:10.1109/TEVC.2015.2459718