Swarm-based search algorithms: A comparative study on sizing optimization of truss structures
Ali Mortazavi · Journal of Structural Engineering & Applied Mechanics · 2023
Metaheuristic algorithms belong to the category of non-deterministic optimization methods. Extensive research conducted in this domain has elucidated that each of these methods possesses distinctive merits and demerits. To illustrate, one algorithm may exhibit a notable penchant for exploration, whereas another algorithm may demonstrate exceptional prowess in exploitation. The judicious selection of a suitable and efficient algorithm for a given problem can profoundly influence both the rate of convergence and the level of accuracy. Over the past few decades, various swarm-based metaheuristic algorithms have been introduced in technical literature. Consequently, undertaking a comprehensive comparative evaluation of a subset of these methodologies can furnish researchers with an essential framework to discern the most appropriate algorithm for their objectives. The current investigation focuses on evaluating and comparing four swarm-based metaheuristic algorithms especially on solving size optimization of truss structures for this purpose. To cover the research conducted in the past two decades as comprehensively as possible, a hierarchical selection process was employed for choosing the methods. As a result, the following algorithms were chosen Firefly Algorithm (FA), Drosophila Food-search Algorithm (DFO), Harris Hawk Optimization (HHO), and Butterfly Optimization Algorithm (BOA). Various characteristics of the selected algorithms, such as convergence rate, diversity variation, complexity, and accuracy of the final solutions, were compared. The findings indicate Harris Hawk Optimization (HHO) could nearly outperform the other selected methods on solving structural size optimization problems.