Year: 2023
Type: article
Abstract: In this paper, the performance of the Differential Evolution algorithm is evaluated when solving real-world problems. A Set of 13 engineering optimisation problems was selected from the fields of mechanics and industry to illustrate the usability of the Differential Evolution algorithm. Twelve variants of the standard Differential Evolution with various settings of the control parameters are compared with 19 state-of-the-art adaptive variants of this algorithm. The results are analysed statistically to achieve significant differences. Three variants of adaptive Differential Evolution provided better results compared to other algorithms. Some adaptive variants of Differential Evolution perform significantly worse than the original Differential Evolution with the fixed setting of the control parameters.
Source: MENDEL
Authors: Petr Bujok, Martin Lacko, Patrik Kolenovský
Institution: University of Ostrava
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $431
OpenAlex ID: W4384565411