A coevolution algorithm based on two-staged strategy for constrained multi-objective problems

Constrained Multiobjective Problem (CMOP) is widely used in engineering applications, but the current constrained Multiobjective Optimization algorithms (CMOEA) often fails to effectively balance convergence and diversity. For this purpose, a two-stage co-evolution constrained multi-objective optimi...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Applied intelligence (Dordrecht, Netherlands) Ročník 52; číslo 15; s. 17954 - 17973
Hlavní autori: Fan, Chaodong, Wang, Jiawei, Xiao, Leyi, Cheng, Fanyong, Ai, Zhaoyang, Zeng, Zhenhuan
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: New York Springer US 01.12.2022
Springer Nature B.V
Predmet:
Artificial Intelligence
Computer Science
Convergence
Evolutionary algorithms
Machines
Manufacturing
Mechanical Engineering
Multiple objective analysis
Optimization
Pareto optimization
Population distribution
Processes
Search process
Constrained multi-objective evolutionary algorithms
Coevolution
Multiobjective optimization
Constraint handling technique
ISSN:0924-669X, 1573-7497
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Constrained Multiobjective Problem (CMOP) is widely used in engineering applications, but the current constrained Multiobjective Optimization algorithms (CMOEA) often fails to effectively balance convergence and diversity. For this purpose, a two-stage co-evolution constrained multi-objective optimization evolutionary algorithm (TSC-CMOEA) is presented to solve constrained multi-objective optimization problems. This method divides the search process into two phases: in the first stage, the synchronous co-evolution is used, and the population corresponding to the help problem and the population corresponding to the raw problem cooperate with each other and share the offspring to produce better solutions, so as to quickly cross the infeasible region and approach the Pareto front; The second stage discards the help problem when it fails and maintains only the evolution of the main population to save computing resources and enhance convergence. The combination of synchronous co-evolution and staged strategy allows the population to traverse infeasible regions more efficiently and converge quickly to feasible and non-dominant regions. The test results on benchmark CMOPs show that the convergence and population distribution of TSC-CMOEA is significantly better than those of NSGA-II, NSGA-III, C-MOEA/D, PPS, ToP and CCMO.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0924-669X
1573-7497
DOI:10.1007/s10489-022-03421-7