• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2026, Vol. 62 ›› Issue (15): 195-220.doi: 10.3901/JME.260619

• 特邀专栏:大模型时代的智能设计与制造 • 上一篇    

扫码分享

大口径反射镜结构优化设计综述及案例

郭冠辰1,2, 梁朋伟3, 庞勇1,2, 李艳婷1,2, 宋学官1,2   

  1. 1. 大连理工大学机械工程学院 大连 116024;
    2. 大连理工大学高性能精密制造全国重点实验室 大连 116024;
    3. 中国工程物理研究院激光聚变研究中心 绵阳 621900
  • 收稿日期:2025-08-20 修回日期:2026-01-16 发布日期:2026-10-13
  • 作者简介:郭冠辰,男,1998年出生。主要研究方向为多学科耦合建模与优化设计。E-mail:ggc1223@mail.dlut.edu.cn;宋学官(通信作者),男,1982年出生,博士,教授,博士研究生导师。主要研究方向为多学科耦合建模与优化设计、装备数字孪生与智能化。E-mail:sxg@dlut.edu.cn
  • 基金资助:
    国家重点研发计划(2023YFB3406900)和国家自然科学基金(52425509,52505247)资助项目。

Review and Case Study of Structural Optimization Design of Large-aperture Mirrors

GUO Guanchen1,2, LIANG Pengwei3, PANG Yong1,2, LI Yanting1,2, SONG Xueguan1,2   

  1. 1. School of Mechanical Engineering, Dalian University of Technology, Dalian 116024;
    2. State Key Laboratory High-Performance Precision Manufacturing, Dalian University of Technology, Dalian 116024;
    3. Laser Fusion Research Center, China Academy of Engineering Physics, Mianyang 621900
  • Received:2025-08-20 Revised:2026-01-16 Published:2026-10-13

摘要: 随着对提高大口径反射镜性能需求的不断提升,如何高效地在多物理场耦合效应下实现反射镜系统多组件协同优化变得至关重要。围绕大口径反射镜及其支撑结构的优化设计展开综述。首先,回顾了大口径反射镜组件的现有设计和优化分析方法,包括一阶设计阶段、拓扑优化、敏感度分析、参数优化和传统代理模型优化方法,并总结当前反射镜及其支撑结构优化设计框架。其次,应用代理模型辅助优化技术补足现有框架在解决高维协同优化问题上的缺陷。最后,通过一个工程示例展示了代理模型辅助优化在反射镜及其支撑组件高维多目标问题优化设计上的有效性。优化结果表明,相较于一阶设计,该方法可以实现双温工况45个参数协同优化,RMS值分别下降73.32%和82.28%;相较于传统优化方法,该方法将优化时间减少两个数量级;相较于传统代理模型优化方法,该方法需要更少的训练数据,并能够快速收敛到满足设计要求的Pareto解集,拥有更好的HV指标并保证解的丰富性、准确性。为现代光机系统设计与分析提供了新的思路和方法,具有重要的工程应用价值。

关键词: 大口径反射镜, 支撑组件, 设计优化, 数据驱动, 代理模型

Abstract: With the increasing demand for improving the performance of large-aperture mirrors, it becomes crucial to efficiently achieve collaborative optimization of multiple components of the mirror system under the coupling effect of multiple physical fields. The optimization design of large-aperture mirrors and their supporting structures is reviewed in this study. First, the existing design and optimization analysis methods of large-aperture mirror components are reviewed, including the first-order design stage, topology optimization, sensitivity analysis, parameter optimization, and traditional surrogate model optimization methods, with a summary provided on the current optimization design framework for mirrors and their supporting structures. Secondly, the surrogate model-assisted optimization technology is applied to supplement the defects of the existing framework in solving high-dimensional collaborative optimization problems. Finally, an engineering example is used to demonstrate the effectiveness of surrogate model-assisted optimization in the optimization design of high-dimensional multi-objective problems of mirrors and their supporting components. The optimization results show that compared with the first-order design, this method can achieve the coordinated optimization of 45 parameters under dual-temperature conditions, and the RMS values are reduced by 73.32% and 82.28% respectively; compared with the traditional optimization method, this method reduces the optimization time by two orders of magnitude; compared with the traditional proxy model optimization method, this method requires less training data and can quickly converge to the Pareto solution set that meets the design requirements, has better HV indicators, and ensures the richness and accuracy of the solution. New ideas and methods are provided for the design and analysis of modern optomechanical systems through this research, which has been demonstrated to hold significant engineering application value.

Key words: large-aperture mirrors, supporting structures, design optimization, data-driven, surrogate model

中图分类号: