中文摘要：

锥形薄壁结构的耐撞性设计过程中,其设计变量和噪声因素都具有一定的波动性,都存在不确定性.传统的优化设计方法由于忽略不确定因素的影响,当设计变量产生波动时,往往会引起设计最优目标超出约束界限或者目标函数对设计变量的波动极为敏感,从而导致设计失效.为了考虑参数的不确定影响,论文提出了一种结合试验设计技术、Kriging近似建模技术、蒙特卡罗模拟技术和6σ质量控制理念的稳健优化设计方法,并应用该方法对泡沫填充锥形薄壁结构的耐撞性进行了优化设计.优化结果表明,该方法提高了锥形薄壁结构的耐撞性,同时也提高了设计变量的可靠性和目标函数的稳健性.由于在优化过程中调用的是Kringing模型,该方法比传统的优化方法具有更高的优化设计效率.算例表明,该方法具有非常强的工程实用性,能大幅提高产品的设计质量.

英文摘要：

Design variables and noise factors have some fluctuations during the process of crashworthiness design for the foam-filled tapered thin-walled structures.Conventional optimization design algorithms ignore the effect of uncertainty factors,which often causes the optimal objective function to go beyond constraints or the objective function to become very sensitive to the functions of the design variables,leading to the design failure.To address the issue,a novel robust optimization based on the experimental design method,Kriging metamodel technique,Monte Carlo method and 6σ method was presented and applied to the crashworthiness design for the foam filled tapered thin walled structures.The results showed that the crashworthiness of the thin-walled structure was improved,and the reliability and robust of the objective function was enhanced.Because the Kriging metamodel was called in the optimization process,the presented method has higher efficiency than the conventional method.Numerical example indicates that this method has a strong engineering practical application and can improve the product quantity significantly.