.png){kind=icon}
.png){kind=icon}
Return
Optimizing mechanical component design with population-based metaheuristics
.png){kind=icon}
PRE
AI
DOI：10.1515/eng-2025-0166.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This study presents a comprehensive comparison of two population-based metaheuristic algorithms, differential evolution (DE) and particle swarm optimization (PSO), applied to the optimization of mechanical component design under both single-objective and multi-objective formulations. The study's novelty lies in its unified evaluation framework, which benchmarks DE and PSO across a set of standardized mechanical design problems and further extends to multi-objective variants (MODE and MOPSO) for full Pareto front analysis. In the single-objective case, both algorithms achieved solutions equal to or better than the best-known results in the literature. PSO generally provided superior solutions and faster convergence, while DE exhibited advantages in computational efficiency and solution stability. In the multi-objective setting, MODE produced broader and more diverse Pareto fronts with higher hypervolume scores, whereas MOPSO delivered competitive results but tended to concentrate around central trade-offs. Importantly, this study provides one of the first systematic comparisons of convergence behavior in single-objective optimization and Pareto front quality through hypervolume analysis in multi-objective mechanical design problems, offering a deeper understanding of convergence-diversity performance across both optimization settings. The findings highlight PSO's strength in rapid exploration and DE's robustness in broad trade-off discovery, providing practical guidance for selecting metaheuristic algorithms in engineering design optimization.
Keywords:
mechanical design
engineering optimization
metaheuristics
particle swarm optimization
differential evolution
industry innovation and infrastructure
Journal
O
IF:
1.5
Papers:
45
Citations:
1.4K