Multi-strategy dynamic multi-objective evolutionary algorithm with hybrid environmental change responses

A key issue in evolutionary algorithms for dynamic multi-objective optimization problems (DMOPs) is how to detect and response environmental changes. Most existing evolutionary algorithms use a single strategy for this purpose. However, single strategy is not always effective. In this paper, we prop...

Full description

Saved in:

Bibliographic Details
Published in:	Swarm and evolutionary computation Vol. 82; p. 101356
Main Authors:	Peng, Hu, Mei, Changrong, Zhang, Sixiang, Luo, Zhongtian, Zhang, Qingfu, Wu, Zhijian
Format:	Journal Article
Language:	English
Published:	Elsevier B.V 01.10.2023
Subjects:	Decomposition Dynamic multi-objective optimization Hybrid environmental change response mechanism Multi-strategy evolutionary operator Multi-strategy evolutionary operator Decomposition Hybrid environmental change response mechanism Dynamic multi-objective optimization
ISSN:	2210-6502
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	A key issue in evolutionary algorithms for dynamic multi-objective optimization problems (DMOPs) is how to detect and response environmental changes. Most existing evolutionary algorithms use a single strategy for this purpose. However, single strategy is not always effective. In this paper, we propose a multi-strategy dynamic multi-objective evolutionary algorithm with hybrid change response (MDMEA-HCR) to solve DMOPs. Our proposed algorithm not only provides a new way for handling dynamics in DMOPs, but also introduce a static multi-objective optimizer based on a multi-strategy evolutionary operator. More specifically, we propose a hybrid environmental change response mechanism to integrate several strategies for prediction and response adjustments. When the environment changes, the hybrid environmental change response strategy makes an initial response to the change, and then the response adjustment mechanism improves the quality of the response population and adjusts its optimization direction to achieve fast tracking of Pareto optimal sets and Pareto optimal fronts in the new environment. During the static optimal optimization phase, a variable neighbor-based multi-strategy evolutionary operator is used to generate new solutions, it is very helpful for both convergence and diversity preservation. MDMEA-HCR has been compared with some other advanced DMOEAs on 31 test instances. Experimental results show that MDMEA-HCR performs better than others on most instances.
ISSN:	2210-6502
DOI:	10.1016/j.swevo.2023.101356