{"title":"A Generational Difference Vector based Tri-Entropy Structure Optimizer for large-scale multiobjective optimization","authors":"Yuhan Xu, Yu Zhang, Wang Hu","doi":"10.1016/j.swevo.2025.102079","DOIUrl":null,"url":null,"abstract":"<div><div>The increasing complexity of large-scale multiobjective optimization problems in engineering and scientific fields in recent years has imposed higher demands on the computational efficiency of algorithms. This paper introduces a novel algorithm named the Generational Difference Vector based Tri-Entropy Structure Optimizer (GDVTSO), which is designed for the efficient execution of large-scale multi-objective optimization tasks. The core idea is to determine a more effective search direction by calculating the local information entropy within the decision space and analyzing the changes in clusters before and after iterations. To this end, the Tri-Entropy Structure Optimizer (TSO) has been designed to more efficiently utilize information entropy for vector updates. Furthermore, the Generational Difference Vector (GDV) mechanism is introduced to provide guidance on search direction for vectors within each cluster. The GDVTSO algorithm demonstrates exceptional compatibility and extensive application potential. In this study, GDVTSO is integrated with two established large-scale optimization techniques, and a hybrid algorithm designated as GDVTSF is proposed through this methodological fusion. Moreover, GDVTSF’s performance exhibits a lower sensitivity to the dimensionality of optimization problems. Experimental results on standard large-scale multiobjective optimization benchmarks demonstrate that GDVTSF outperforms the current state-of-the-art optimization algorithms. Furthermore, it remarkably maintains its superior performance even when applied to high-dimensional problems with up to 10,000 decision variables.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"98 ","pages":"Article 102079"},"PeriodicalIF":8.5000,"publicationDate":"2025-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Swarm and Evolutionary Computation","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2210650225002378","RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0
Abstract
The increasing complexity of large-scale multiobjective optimization problems in engineering and scientific fields in recent years has imposed higher demands on the computational efficiency of algorithms. This paper introduces a novel algorithm named the Generational Difference Vector based Tri-Entropy Structure Optimizer (GDVTSO), which is designed for the efficient execution of large-scale multi-objective optimization tasks. The core idea is to determine a more effective search direction by calculating the local information entropy within the decision space and analyzing the changes in clusters before and after iterations. To this end, the Tri-Entropy Structure Optimizer (TSO) has been designed to more efficiently utilize information entropy for vector updates. Furthermore, the Generational Difference Vector (GDV) mechanism is introduced to provide guidance on search direction for vectors within each cluster. The GDVTSO algorithm demonstrates exceptional compatibility and extensive application potential. In this study, GDVTSO is integrated with two established large-scale optimization techniques, and a hybrid algorithm designated as GDVTSF is proposed through this methodological fusion. Moreover, GDVTSF’s performance exhibits a lower sensitivity to the dimensionality of optimization problems. Experimental results on standard large-scale multiobjective optimization benchmarks demonstrate that GDVTSF outperforms the current state-of-the-art optimization algorithms. Furthermore, it remarkably maintains its superior performance even when applied to high-dimensional problems with up to 10,000 decision variables.
期刊介绍:
Swarm and Evolutionary Computation is a pioneering peer-reviewed journal focused on the latest research and advancements in nature-inspired intelligent computation using swarm and evolutionary algorithms. It covers theoretical, experimental, and practical aspects of these paradigms and their hybrids, promoting interdisciplinary research. The journal prioritizes the publication of high-quality, original articles that push the boundaries of evolutionary computation and swarm intelligence. Additionally, it welcomes survey papers on current topics and novel applications. Topics of interest include but are not limited to: Genetic Algorithms, and Genetic Programming, Evolution Strategies, and Evolutionary Programming, Differential Evolution, Artificial Immune Systems, Particle Swarms, Ant Colony, Bacterial Foraging, Artificial Bees, Fireflies Algorithm, Harmony Search, Artificial Life, Digital Organisms, Estimation of Distribution Algorithms, Stochastic Diffusion Search, Quantum Computing, Nano Computing, Membrane Computing, Human-centric Computing, Hybridization of Algorithms, Memetic Computing, Autonomic Computing, Self-organizing systems, Combinatorial, Discrete, Binary, Constrained, Multi-objective, Multi-modal, Dynamic, and Large-scale Optimization.