复合材料科学与工程 ›› 2021, Vol. 0 ›› Issue (3): 14-20.DOI: 10.19936/j.cnki.2096-8000.20210328.002

• 基础研究 • 上一篇    下一篇

基于改进差分进化算法的复合材料结构优化设计

马森1, 赵启林2, 施丽铭3, 单成之1   

  1. 1.武警警官学院 训练基地,广州510440;
    2.南京工业大学 机械与动力工程学院,南京211816;
    3.北京空间飞行器总体设计部,北京100094
  • 收稿日期:2020-07-06 出版日期:2021-03-28 发布日期:2021-04-30
  • 作者简介:马森(1989-),男,博士,讲师,主要从事复合材料结构力学、复合材料结构优化设计方面的研究,mesir1989@sina.com。
  • 基金资助:
    中国载人航天领域预先研究项目(载人四批040101);国家重点研发计划项目(2017YFC0405103)

OPTIMIZATION OF COMPOSITES STRUCTURE BASED ON IMPROVED DIFFERENTIAL EVOLUTION ALGORITHM

MA Sen1, ZHAO Qi-lin2, SHI Li-ming3, SHAN Cheng-zhi1   

  1. 1. Training Base, Officers College of PAP, Guangzhou 510440, China;
    2. School of Mechanical and Power Engineering, Nanjing Tech University, Nanjing 211816, China;
    3. Beijing Institute of Spacecraft System Engineering, Beijing 100094, China
  • Received:2020-07-06 Online:2021-03-28 Published:2021-04-30

摘要: 提出了一种改进的S形变异差分进化算法并对该算法在复合材料结构优化方面的应用进行了研究。首先,分析了控制参数对差分进化算法收敛性和寻优能力的影响机理;其次,针对控制参数对差分进化算法的影响机理,借鉴生物繁殖的“S形曲线”规律,提出变异率的“S形曲线”变化趋势,提出改进的差分进化算法SDE,数值算例证明该算法具有较好的全局寻优能力;最后,针对复合材料结构优化设计问题,建立了基于SDE算法的优化计算系统,利用该优化计算系统对某空间运载器复合材料构件进行了优化设计,优化结果较初始设计质量减小36.5%。

关键词: 复合材料结构, 优化设计, 差分进化算法, S形曲线

Abstract: An improved S-shaped mutation Differential Evolution algorithm is proposed and its application in optimization of composites structure is studied. Firstly, the influence mechanism of control parameters on the convergence and optimization ability of the Differential Evolution algorithm is analyzed. Secondly, aiming at the influence mechanism and learning from the "S-shaped curve" law of biological reproduction, a new change trend of the mutation rate is proposed, and an improved Differential Evolution algorithm (SDE) is proposed, advantage of the improved algorithm is proved by numerical examples. Finally, an optimization computing system based on the SDE algorithm for composites structure is established, which is used to optimize a composites component of a space vehicle, and the weight is 36.5% lower than the initial design.

Key words: composites structure, optimization, the differential evolution algorithm, S-shaped curve

中图分类号: