中文摘要：

为获得不同运行速度和工况下的高速列车车厢侧墙结构，在拓扑优化结构基础上进行了多目标优化研究.将侧墙夹层板质量、柔度、最大变形作为优化目标，侧墙五段夹层结构的面板和夹心厚度为变量，车厢气压变化梯度为约束函数，利用代理模型技术，建立了各目标、约束函数与变量之间的代理模型，通过非支配遗传算法 NSGA-II，得到多目标的 Pareto 解集.该 Pareto 解集中的夹层板结构比拓扑优化得到的夹层板结构的性能提高了 8.21%到 33.58%，设计时可根据具体的要求和经验从 Pareto 解集中进行选择，从而为不同运行速度和工况下的高速列车车厢断面结构设计提供了多种选择方案.

英文摘要：

To investigate the structures of high speed train side walls applicable to different running speeds and operation conditions,a multi-objective optimization design is carried out following structure topology optimization.The weight of sandwich plate,static compliance and maximum deformation are defined as the objectives,the thickness of face panels and cores in five parts of the side wall as the variables while the changing air pressure gradient in compartments as the constraint.The surrogate model techniques are implemented for constructing the response surfaces of objective and constraint functions.Then a multi-objective optimization is performed with NSGA-II to generate a Pareto solution set.The structure performance in Pareto set is greatly improved by 8.21% to 33.58% than that from topology structure,besides,the Pareto solution set provides many alternative Pareto-optimal solutions for optimization design of the sandwich plate section in high-speed trains.