Multi-strategy Improved Seagull Optimization Algorithm

Aiming at the shortcomings of seagull optimization algorithm in the process of searching for optimization, such as slow convergence speed, low precision, easy falling into local optimal, and performance dependent on the selection of parameters, this paper proposes an improved gull optimization algor...

Full description

Saved in:
Bibliographic Details
Published in:International journal of computational intelligence systems 2023-09, Vol.16 (1), p.1-27, Article 154
Main Authors: Li, Yancang, Li, Weizhi, Yuan, Qiuyu, Shi, Huawang, Han, Muxuan
Format: Article
Language:English
Subjects:
Algorithm improvement
Artificial Intelligence
Computational Intelligence
Control
Engineering
Mathematical Logic and Foundations
Mechatronics
Research Article
Robotics
Seagull optimization algorithm
Simulation experiment
Structural optimization
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Aiming at the shortcomings of seagull optimization algorithm in the process of searching for optimization, such as slow convergence speed, low precision, easy falling into local optimal, and performance dependent on the selection of parameters, this paper proposes an improved gull optimization algorithm based on multi-strategy fusion based on the analysis of gull population characteristics. Firstly, L–C cascade chaotic mapping is used to initialize the population so that seagulls are more evenly distributed in the initial solution space. Secondly, to improve the algorithm’s global exploration ability in the early stage, the nonlinear convergence factor is incorporated to adjust the position of seagulls in the migration stage. At the same time, the group learning strategy was introduced after the population position update to improve the population quality and optimization accuracy further. Finally, in the late stage of the algorithm, the golden sine strategy of the Levy flight guidance mechanism is used to update the population position to improve the population’s diversity and enhance the local development ability of the algorithm in the late stage. To verify the optimization performance of the improved algorithm, CEC2017 and CEC2022 test suites are selected for simulation experiments, and box graphs are drawn. The test results show that the proposed algorithm has apparent convergence speed, accuracy, and stability advantages. The engineering case results demonstrate the proposed algorithm’s advantages in solving complex problems with unknown search spaces.
ISSN:1875-6883
1875-6883
DOI:10.1007/s44196-023-00336-0