A coevolutionary algorithm using Self-organizing map approach for multimodal multi-objective optimization
Multimodal Multi-Objective Problems (MMOPs) are frequently encountered in the real world. Traditional Multimodal Multi-Objective Evolutionary Algorithms (MMEAs) often find multiple Pareto optimal solutions with the same objective values. However, in real-world problems, there often exist multiple gl...
Saved in:
Published in | Applied soft computing Vol. 164; p. 111954 |
---|---|
Main Authors | , , , , , |
Format | Journal Article |
Language | English |
Published |
Elsevier B.V
01.10.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Multimodal Multi-Objective Problems (MMOPs) are frequently encountered in the real world. Traditional Multimodal Multi-Objective Evolutionary Algorithms (MMEAs) often find multiple Pareto optimal solutions with the same objective values. However, in real-world problems, there often exist multiple global optimal solutions and local optimal solutions at the same time. Ensuring that these solution sets are obtained simultaneously is the concern of most current researchers. To address this issue, this paper proposes a novel multimodal multi-objective evolutionary algorithm named CoSOMEA. In the CoSOMEA, a Self-organizing map (SOM) neural network is used to extract the information of decision space to ensure better exploration of the global optima and exploitation of the local optima. Meanwhile, coevolutionary mechanism are used to ensure a balance between the exploration and exploitation in order to avoid the algorithm falling into local areas. The three test suites named IDMP, IDMP_ee and MMF are adopted to verify the effectiveness of proposed algorithm. Experimental results demonstrate that the CoSOMEA exhibits competitive performance in solving MMOPs compared to other state-of-the-art MMEAs.
•An effective multimodal multi-objective optimization algorithm is designed.•Applying SOM neural networks to Multimodal Multi-Objective Evolutionary Algorithms.•A Coevolutionary mechanism is designed to guide populations away from falling into local optimal regions.•The exploration of global regions and exploitation of local regions in decision space are enhanced. |
---|---|
ISSN: | 1568-4946 |
DOI: | 10.1016/j.asoc.2024.111954 |