基于自适应混合代理模型的功能梯度泡沫填充多胞结构的耐撞性优化设计理论研究

项目名称： 基于自适应混合代理模型的功能梯度泡沫填充多胞结构的耐撞性优化设计理论研究

项目编号： No.11302075

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 数理科学和化学

项目作者： 尹汉锋

作者单位： 湖南大学

项目金额： 23万元

中文摘要： 功能梯度泡沫填充多胞结构以其良好的吸能特性在汽车、列车、航空航天等领域具有非常好的应用前景。该结构相对普通泡沫填充结构更为复杂，其耐撞性受多胞结构的材料、尺寸、形状和泡沫的梯度分布规律、密度等多种因素的影响。另外，该结构耐撞性设计所涉及到的接触碰撞问题本身就是一个复杂的强非线性问题。所以，在该结构的耐撞性优化中，通常存在计算精度和计算效率低的问题，难以满足工程设计要求。本项目将结合耐撞性理论、动态采样方法、混合代理模型构建方法和多目标粒子群算法，发展基于自适应混合代理模型的结构耐撞性多目标优化设计方法，以解决采用传统方法对功能梯度泡沫填充多胞结构进行耐撞性优化设计时出现的精度和效率低的问题。此外，相关理论和方法还将应用到我们提出的一类新型功能梯度泡沫填充多胞结构的耐撞性优化设计中，以期满足航天着陆器缓冲机构的耐撞性和轻量化的苛刻要求。

中文关键词： 耐撞性；优化设计；代理模型；多胞结构；泡沫填充结构

英文摘要： Functionally graded foam-filled multicell structure (FGFMS) has good application prospects in the fields of vehicle, train, aerospace and et al. due to its excellent energy absorption capacity. The construction of this structure is more complicated than that of the ordinary foam-filled structure. The crashworthiness of this structure is affected by lots of factors such as the material, size, shape of the multicell structure and the grading pattern, densities of the filled foams and et al. In addition, the contact-impact problem itself involved in the crashworthiness design of this structure is a complicated strong nonlinear problem. Thus, the problems of low numerical accuracy and low numerical efficiency are likely to appear in the crashworthiness optimization design of this structure. Due to these problems, it is difficult to meet the demand of the engineering design. In order to deal with the problem of low numerical accuracy and low numerical efficiency when carrying out the crashworthiness optimization design for the FGFMS with the traditional method, the crashworthiness theory, the dynamic sampling method, the ensemble surrogate modeling method and the multiobjective particle swarm optimization (MOPSO) algorithm are jointly employed to develop a novel multiobjective crashworthiness optimization design meth

英文关键词： Crashworthiness；Optimization design；Metamodel；Multi-cell structure；Foam-filled structure