An Improved Secretary Bird Optimization Algorithm
Hongwu Qin, Songhao Yang, Zhenqi Liu, Guangxi Li · 2024
The traditional secretary bird optimization algorithm is known to exhibit certain deficiencies, including a slow convergence speed and a tendency to fall into local optimization. This paper presents an enhanced version of the algorithm, addressing these limitations. The proposed algorithm incorporates three key innovations. First, it employs Circle chaotic mapping to generate an initialized population with greater biodiversity. Second, it integrates a differential evolution strategy into the iterative process, balancing global search and local optimization to improve convergence accuracy. Additionally, a population diversity strategy is introduced to enhance convergence speed and optimization accuracy. This strategy was tested on six benchmark functions and compared with other optimization algorithms. Experimental results demonstrate that ISBOA achieves optimal convergence speed and optimization accuracy, validating its effectiveness on complex problems.