An elite exemplar learning-based artificial electric field algorithm for numerical optimization
Physica Scripta · 03.04.2026 · 101 (13) · 135202
DOI: 10.1088/1402-4896/ae557c ↗Kurz gefasst
Fachartikel in der Zeitschrift Physica Scripta, online erschienen am 30.03.2026. Verfasst von Qiuhong Lin und 1 weiteren Autor:innen. Die Arbeit nennt 65 Literaturangaben.
Abstract (Originalsprache)
As a new stochastic optimizer, artificial electric field algorithm (AEFA) has been confirmed to be effective in tackling diverse real-world problems. Nevertheless, AEFA works with a simple search mechanism and a fixed attenuation factor, which makes it prone to stagnation and premature convergence when handling complex search tasks. In response to these issues, an elite exemplar learning-based AEFA (EL-AEFA) is proposed. More specifically, in EL-AEFA, elite individuals with distinct merits are stored in different exemplar pools. Then, an elite exemplar learning strategy is developed, which makes multiple individuals saved in the exemplar pool co-direct the entire population in each generation. Further, a time-varying parameter control scheme is designed to dynamically generate a more suitable attenuation factor. Numerical experiments are executed on 42 benchmark functions to assess the performance of EL-AEFA. Moreover, three chaotic time series prediction problems are employed to examine the effectiveness and versatility of EL-AEFA. Experimental results reveal that the presented EL-AEFA is superior to the original AEFA, four recently developed variants of AEFA, and 35 powerful algorithms. The source code of EL-AEFA is publicly available at https://github.com/lqhgxgl/EL-AEFA.
Zitieren
APA
Lin, Q., & Lu, Q. (2026). An elite exemplar learning-based artificial electric field algorithm for numerical optimization. Physica Scripta, 101(13), 135202. https://doi.org/10.1088/1402-4896/ae557c
Vancouver
Lin Q, Lu Q. An elite exemplar learning-based artificial electric field algorithm for numerical optimization. Physica Scripta. 2026;101(13):135202. doi:10.1088/1402-4896/ae557c
BibTeX
@article{Lin2026,
author = {Lin, Qiuhong and Lu, Qiuli},
title = {An elite exemplar learning-based artificial electric field algorithm for numerical optimization},
journal = {Physica Scripta},
year = {2026},
volume = {101},
number = {13},
pages = {135202},
doi = {10.1088/1402-4896/ae557c}
}
EUVIMED