数据驱动的结构拓扑优化技术综述  

Review of data-driven structural topology optimization technology

作　　者：白文灿 刘莉[1] 田维勇 Bai Wencan;Liu Li;Tian Weiyong(Beijing Institute of Technology,Beijing 100081,China)

机构地区：[1]北京理工大学,北京100081

出　　处：《战术导弹技术》2023年第6期1-12,33,共13页Tactical Missile Technology

摘　　要：数据驱动的结构拓扑优化通过学习大量结构拓扑优化设计过程来建立拓扑优化模型,近年来备受关注。综述了结构拓扑优化技术的发展历程,回顾了结构拓扑优化技术的发展现状和研究进展,将发展历程分为四个阶段,并对每一阶段进行分析和总结;详细介绍了数据驱动的结构拓扑优化技术,按照建模方式的不同将其分类为基于卷积神经网络、支持向量回归、生成对抗网络和仿生算法的拓扑优化技术;结合各方法的研究成果,介绍了数据驱动的结构拓扑优化技术在工程设计中的应用,并对其未来的发展方向进行了展望;指出了数据驱动的结构拓扑优化技术面临的挑战和未来需要解决的问题,为拓扑优化领域未来的发展提供参考和借鉴。Data-driven structural topology optimization has attracted much attention recently,which achieving structural topology optimization by learning a large number of structural topology optimization design processes to establish a topology optimization model.It is summarized the development process of structural topology optimization and reviewed the development status and research progress of structural topology optimization.The development process is divided into four stages,and each stage is analyzed and summarized.The data-driven structural topology optimization technology is introduced in details.According to the establishment of different models,the topology optimization technology is classified into convolutional neural network,support vector regression,generative adversarial network and bionic algorithm.Combined with the research results of each method,the application of data-driven structure topology optimization technology in engineering design is introduced,and the future development tendency is prospected.The challenges and problems to be solved in the future of data-driven structural topology optimization are pointed out,which can provide reference for the future development of structural topology optimization.

关 键 词：拓扑优化 机器学习 卷积神经网络 支撑向量回归 生成对抗网络 仿生算法 数据驱动 

分 类 号：TJ03[兵器科学与技术—兵器发射理论与技术]