Pied Kingfisher optimizer with mixed strategy improvement and its application in robot path planning
Year: 2025
Type: preprint
Abstract: Optimization algorithms find broad applications in fields such as machine learning, engineering design, and control. The Pied Kingfisher Optimizer, emerging in recent years, is an efficient metaheuristic algorithm that draws inspiration from the distinctive hunting strategies and symbiotic interactions of pied kingfisher observed in their natural environment. However, it still grapples with sluggish convergence and a tendency to become trapped in local optima when tackling specific complex problems. To counter these drawbacks, this study develops a modified Pied Kingfisher Optimizer. This improvement integrates three hybrid mechanisms: Firefly Algorithm, Differential Evolution, and an adaptive t-distribution mutation operator. Ablation testing via the CEC2017 and CEC2022 suites evaluated the proposed algorithm’s operational capability. Furthermore, the upgraded algorithm was rigorously benchmarked against several other metaheuristic optimizers. Empirical findings indicate that the refined Pied Kingfisher Optimizer attains a higher overall performance level compared to the chosen competitor algorithms. This performance superiority was statistically confirmed through the Friedman test. Additionally, to investigate the practical utility of the enhanced Pied Kingfisher Optimizer for control applications, it was tested on robot path planning tasks, where it demonstrated competitive performance. These results further underscore the enhanced algorithm’s efficacy within the control domain.
Authors: XiangNan Xiong, Peng Yu, Xue-Bo Chen
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4413909079