基于稳健优化的轿车车身轻量化设计方法研究

项目名称： 基于稳健优化的轿车车身轻量化设计方法研究

项目编号： No.50875164

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 能源与动力工程

项目作者： 朱平

作者单位： 上海交通大学

项目金额： 31万元

中文摘要： 减轻汽车质量以实现节能减排是汽车工业发展的核心问题，新能源汽车与传统汽车都迫切需要轻量化设计与改进。轿车车身轻量化设计必须满足耐撞安全、NVH、疲劳等性能，是一个多学科、多参数的优化问题。实际中，车身零件的板厚、截面尺寸及材料参数与其名义值存在一定的变差，直接影响车身结构性能，降低了轻量化设计方案的工程可靠性。本项目利用数值仿真与虚拟试验技术，研究轿车车身各项结构性能指标的响应面模型的建模方法并建立模型，确定了各性能指标与设计变量（如板厚、材料参数）及随机噪声因素（如材料参数）的定量关系。通过对设计变量及噪声因素的随机变量统计特性的研究，揭示了性能指标响应面模型的随机函数统计特性，提出综合考虑设计变量及噪声因素变差的车身轻量化稳健设计准则函数，建立了多参数、多约束的车身轻量化设计数学模型，形成了基于稳健优化的车身轻量化设计方法，并开发了软件工具，促进我国汽车轻量化设计能力的提高。

中文关键词： 车身轻量化；结构性能；稳健优化；响应面法；试验设计

英文摘要： Lightening automobile is the kernel of development faced to the automotive industry to realize energy economy and both the conventional automobile and new energy car require urgently the lightweight design. The lightweight design of autobody structure must meet the requirements of structural performances such as crashworthiness, NVH, fatigue and so on, which is in essence a multidisciplinary optimization problem. However, the lightweight design without considerating the variation of gauge thickness and material mechanics parameter, caused by the uncertainties in the real engineering application, will lead to the great loss of reliability and feasibility. In this project, complex metamodeling techniques, including support vector regression (SVR), moving least square (MLS), kriging (KG), radial basis function (RBF) and artificial neural net work (ANN), investigated for approximating as accurate and efficient as possible the function relationship of the strctural performance indicator and the design variable (thickness, mechanics parameter of material) and random noise factor (mechanics parameter of material). The statistical character of the structural performance indicator is shown according to the probability distribution of design variable and noise factor. Furthermore, the robust and realiability design optimization method is investigated for creating the mathematical model of multiparameter and multiconstraint design optimization and obtaining the robust optimum design considering both the robustness of objectives and the reliability of constraints. Finally, the lightweight design method based on the robust optimization for automotive body structure is established. The lighweight design obtained by using the method will be ensured to be more reliable and feasible in the real engineering application. The study in this project plays a vital role in improving the ability of lightweight design of autobody structure for our domestic automovtive industry.

英文关键词： Lightweight of body structure; structural performance; robust optimization; response surface method; experimental design