学术论文

      地下工程智能连续优化方法的实现Ⅰ:参数化设计及优化方法

      Implementation of the intelligent and continuous optimization method in underground engineering,PartⅠ. Parametric design and optimization method

      摘要:
      地下工程空间分布优化设计是大型地下工程稳定性分析的关键。如今,制约其发展的瓶颈主要有两个:一是模型的建立;二是优化方法的实现。针对以上问题,以大型有限元软件ABAQUS为平台,基于其标准设计语言Python对其前后处理进行二次开发,将参数化模型可动边界的非线性关系编入到程序中,实现了有限元参数化设计;通过MATLAB语言编程,建立基于Nelder-Mead法与精确罚函数法的联合优化方法,并将优化分析程序与有限元参数化设计相结合,实现了地下工程问题的“智能连续”优化。文中涉及到工程应用将在本文的Ⅱ部分给出。
      Abstract:
      The optimization design of underground engineering space distribution is the key of stability anal-ysis in a large-scale underground engineering. Nowadays, there are two main bottlenecks to restrict its de-velopment. One is the problem of model establishing and the other is the implementation of the optimiza-tion method. Taking the general-purpose finite element software ABAQUS as a platform, its standard design language Python is used to make secondary development for the pre-processing and post-processing. The moving boundary nonlinear relationship of parametric model is made into the program to realize the fem-based parametric design. Using MATLAB language for programming,a combined optimization was estab-lished on the basis of Nelder-Mead algorithm and exact penalty function method, and further combining the optimum analysis program with finite element parametric design, the intelligent and continuous optimiza-tion method in underground engineering has been realized. The engineering application involved in this pa-per will be given in Part Ⅱ of this paper series.
      作者: 王辉 [1] 陈卫忠 [2] 李廷春 [3] 王清标 [1]
      Author: WANG Hui [1] CHEN Wei-zhong [2] LI Ting-chun [3] WANG Qing-biao [1]
      作者单位: 山东科技大学 资源与土木工程系,山东 泰安,271019 中国科学院 武汉岩土力学研究所 岩土力学与工程国家重点实验室,湖北 武汉,430071 山东科技大学 山东省土木工程防灾减灾重点实验室,山东 青岛,266510
      刊 名: 水利学报 ISTICEIPKU
      年,卷(期): 2014, (1)
      分类号: TU923
      机标分类号: TP3 G64
      在线出版日期: 2014年2月22日
      基金项目: 山东科技大学人才引进科研启动基金项目资助;山东科技大学泰安校区科研创新团队支持计划;国家自然科学基金资助项目(51279096、41102193、41372289);岩土力学与工程国家重点实验室资助课题