Mehta Pranav, Tejani Ghanshyam G, Mousavirad Seyed Jalaleddin
Department of Mechanical Engineering, Dharmsinh Desai University, Nadiad, Gujarat, 387001, India.
Department of Research Analytics, Saveetha Dental College and Hospitals, Saveetha Institute of Medical and Technical Sciences, Saveetha University, Chennai, 600077, India.
Sci Rep. 2025 Apr 25;15(1):14575. doi: 10.1038/s41598-025-97133-w.
Optimizing a multi-objective structure is a challenging design problem that requires handling several competing goals and constraints. Despite their success in resolving such issues, metaheuristics can be difficult to apply due to their stochastic nature and restrictions. This work proposes the multi-objective crystal structure optimizer (MOCRY), a potent and effective optimizer, to address this problem. The MOCRY algorithm, also known as MOCRY2arc, is built on a two-archive idea centered on diversity and convergence, respectively. The efficacy of MOCRY2arc in solving five typical planar and spatial real-world structure optimization issues was assessed. Because of these problems, safety and size limits were put on discrete cross-sectional regions and component stress. At the same time, different goals were being pursued, such as making nodal points bend more and reducing the mass of trusses. Four recognized standard evaluators-Hypervolume (HV), Generational-Inverted Generational Distance (GD, IGD), Spacing to Extent Metrics (STE), convergence, and diversity plots-were utilized to compare the results with those of nine sophisticated optimization techniques, including MOCRY and NSGA-II. Moreover, the Friedman rank test and comparison analysis showed that MOCRY2arc performed better at resolving big structure optimization issues in a shorter amount of computing time. In addition to identifying and realizing effective Pareto-optimal sets, the recommended method produced strong variety and convergence in the objective and choice spaces. As a result, MOCRY2arc may be a useful tool for handling challenging multi-objective structure optimization issues.
优化多目标结构是一个具有挑战性的设计问题，需要处理多个相互竞争的目标和约束条件。尽管元启发式算法在解决此类问题方面取得了成功，但由于其随机性和局限性，可能难以应用。这项工作提出了多目标晶体结构优化器（MOCRY），一种强大而有效的优化器，以解决这个问题。MOCRY算法，也称为MOCRY2arc，基于分别以多样性和收敛性为中心的双存档思想构建。评估了MOCRY2arc在解决五个典型的平面和空间实际结构优化问题方面的有效性。由于这些问题，对离散横截面区域和部件应力设置了安全和尺寸限制。同时，正在追求不同的目标，例如使节点弯曲更多并减少桁架的质量。使用四个公认的标准评估器——超体积（HV）、世代-反向世代距离（GD、IGD）、间距与范围度量（STE）、收敛性和多样性图——将结果与包括MOCRY和NSGA-II在内的九种先进优化技术的结果进行比较。此外，Friedman秩检验和比较分析表明，MOCRY2arc在更短的计算时间内解决大型结构优化问题方面表现更好。除了识别和实现有效的帕累托最优集外，推荐的方法在目标空间和选择空间中产生了很强的多样性和收敛性。因此，MOCRY2arc可能是处理具有挑战性的多目标结构优化问题的有用工具。
