Year: 2024
Type: article
Abstract: As an efficient meta-heuristic technique, artificial electric field algorithm (AEFA) has been extensively applied to tackle various challenging tasks posed by practical scenarios. However, in the classical AEFA, the fitness function has a cumulative effect on the charge, resulting in limited search capability. To address this issue, a modified AEFA (MAEFA) is presented in this paper. More specifically, a novel charge calculation scheme is introduced to overcome the cumulative effect by gradually distinguishing the charges of particles during the evolutionary process. Further, an alternating search strategy is developed to calculate the total electrostatic force, thereby reinforcing the guiding effect of excellent individuals on the entire population. Subsequently, the performance of MAEFA is investigated using 42 well-benchmarked functions, two chaotic time series prediction problems, and two engineering design problems. Experimental results reveal that MAEFA is more competitive in comparison with several established AEFAs and 20 popular meta-heuristic techniques.
Source: Physica Scripta
Authors: Qiuhong Lin, Lieping Zhang, Jiatang Cheng
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
Funder: Guilin University of Technology
Award: G5520444791
OpenAlex ID: W4404332047