Catch fish optimization algorithm based on improved multi-strategy fusion

Shi, Panling

doi:10.1088/2631-8695/ae2078

Return

Catch fish optimization algorithm based on improved multi-strategy fusion

2025-12-31

PRE

DOI：10.1088/2631-8695/ae2078

Original

Original request for help

Save

Abstract

En 中文

The Catch Fish Optimization Algorithm (CFOA) is a metaheuristic algorithm that simulates the fishing process of fishermen. It performs well in terms of optimization performance and clustering problems, but it has certain drawbacks in the independent search and collective capture stages,resulting in low exploration efficiency and weak local development capabilities. To solve the aforementioned problems, a Fishing Optimisation Algorithm based on an Improved Multi-strategy Fusion (GCFOA) is proposed. This algorithm adopts a Markov-guided Cauchy perturbation strategy and constructs a dynamic elite selection pool to solve the problem of inefficient exploration caused by the random selection of reference individuals. Meanwhile, it combines the heavy-tailed characteristics of the Cauchy distribution to enhance global exploration and the ability to jump out of local optima. In order to balance the exploration and exploitation processes, a Sigmoid dynamic adjustment mechanism is designed, which replaces the fixed threshold with a smooth transition method to improve adaptability to complex solution spaces. The Broyden-Fletcher-Goldfarb-Shanno ( BFGS ) is introduced for local search to refine the global optimal solution, making up for the deficiency of the original algorithm in local search capability and thereby improving convergence accuracy. The proposed GCFOA algorithm was evaluated on 41 benchmark functions and several practical engineering optimization problems, and its performance was compared with six other swarm intelligence algorithms. Experimental results demonstrate that GCFOA achieves significantly higher convergence accuracy and stability than the competing algorithms. Furthermore, the Friedman and Wilcoxon signed-rank tests statistically confirm the effectiveness of the proposed improvements.

Keywords:

catch fish optimization algorithm

sigmoid adjustment mechanism

BFGS algorithm

engineering problems