.png){kind=icon}
.png){kind=icon}
Return
An improved generalized evolutionary algorithm for constrained multimodal multiobjective optimization
.png){kind=icon}
OA
AI
C
W
J
Y
K
W
DOI：10.1007/s40747-025-02164-5.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
When searching for constrained Pareto solutions in constrained multimodal multiobjective optimization problems (CMMOPs), infeasible solutions and local optimal solutions play an important role. Specifically, infeasible solutions can guide the algorithm toward regions that may eventually lead to feasible solutions or provide insights into the problem’s constraint boundaries. Moreover, local optimal solutions can enhance population diversity. In terms of this situation, this paper proposes a novel constrained multimodal multiobjective algorithm, which introduces a dynamic local fitness evaluation and a self-cleaning mechanism (CMMOGA_DLF). The dynamic local fitness evaluation effectively enhances population diversity by utilizing local optimal solutions while maintaining convergence. In addition, by introducing the local dominance criterion (LDC) and using high-quality infeasible solutions, the algorithm breaks through the infeasibility barrier in the early stage and gradually approaches the feasible area in the later stage. Meanwhile, the self-cleaning mechanism efficiently removes similar solutions and further improves the diversity of the population. Finally, the effectiveness of the proposed method is verified by comparing it with five existing algorithms on 31 test functions and a local selection problem.
Keywords:
Constraints
Multimodal
Multiobjective
Evolutionary algorithms
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.6
Papers:
2.1K
Citations:
6.6K
Organization
Cited Papers