A new approach for solving global optimization and engineering problems based on modified Sea Horse Optimizer
Fatma A Hashim,Reham R Mostafa,Ruba Abu Khurma,Raneem Qaddoura,P A Castillo
DOI: https://doi.org/10.1093/jcde/qwae001
2024-01-04
Journal of Computational Design and Engineering
Abstract:Sea Horse Optimizer (SHO) is a noteworthy metaheuristic algorithm that emulates various intelligent behaviors exhibited by sea horses, encompassing feeding patterns, male reproductive strategies, and intricate movement patterns. To mimic the nuanced locomotion of sea horses, SHO integrates the logarithmic helical equation and Levy flight, effectively incorporating both random movements with substantial step sizes and refined local exploitation. Additionally, the utilization of Brownian motion facilitates a more comprehensive exploration of the search space. This study introduces a robust and high-performance variant of the SHO algorithm named mSHO. The enhancement primarily focuses on bolstering SHO's exploitation capabilities by replacing its original method with an innovative local search strategy encompassing three distinct steps: a neighborhood-based local search, a global non-neighbor-based search, and a method involving circumnavigation of the existing search region. These techniques improve mSHO algorithm's search capabilities, allowing it to navigate the search space and converge toward optimal solutions efficiently. To evaluate the efficacy of the mSHO algorithm, comprehensive assessments are conducted across both the CEC2020 benchmark functions and nine distinct engineering problems. A meticulous comparison is drawn against nine metaheuristic algorithms to validate the achieved outcomes. Statistical tests, including Wilcoxon's rank-sum and Friedman's tests, are aptly applied to discern noteworthy differences among the compared algorithms. Empirical findings consistently underscore the exceptional performance of mSHO across diverse benchmark functions, reinforcing its prowess in solving complex optimization problems. Furthermore, the robustness of mSHO endures even as the dimensions of optimization challenges expand, signifying its unwavering efficacy in navigating complex search spaces. The comprehensive results distinctly establish the supremacy and efficiency of the mSHO method as an exemplary tool for tackling an array of optimization quandaries. The results show that the proposed mSHO algorithm has a total rank of 1 for CEC'2020 test functions. In contrast, the mSHO achieved the best value for the engineering problems, recording a value of 0.012665, 2993.634, 0.01266, 1.724967, 263.8915, 0.032255, 58507.14, 1.339956, and 0.23524 for the pressure vessel design, speed reducer design, tension/compression spring, welded beam design, three-bar truss engineering design, industrial refrigeration system, multi-Product batch plant, cantilever beam problem, multiple disc clutch brake problems, respectively. Source codes of mSHO are publicly available at https://www.mathworks.com/matlabcentral/fileexchange/135882-improved-sea-horse-algorithm.
computer science, interdisciplinary applications,engineering, multidisciplinary