基于DOE参数筛选的SPH鸟体参数反演  

Parameter Inversion of SPH Bird Model Based on DOE Parameter Selection

作　　者：罗军[1] 刘长虹[1] 洪清泉 鞠锋[1] 丁敏[1] 

机构地区：[1]华东理工大学机械与动力工程学院承压系统与安全教育部重点实验室,上海200237 [2]Altair中国软件有限公司,上海200082

出　　处：《机械设计与制造》2012年第9期90-92,共3页Machinery Design & Manufacture

基　　金：国家“863”计划资助(2006AA042425);教育部新世纪优秀人才计划(NCET.06.0414);上海市重点学科建设项目(B503)

摘　　要：鸟体参数的准确性对鸟撞仿真精度有重大影响,参数反演可以克服人工试凑法的局限性,搜寻出合理的鸟体参数,提高鸟撞仿真精度。在HyperStudy多学科优化平台下,先采用正交实验设计方法挑选出对位移结果敏感的鸟体参数,以简化优化问题的复杂性和计算量,然后采用自适应响应面法进行鸟体的参数反演的多目标优化,优化目标为最小化鸟撞位置处仿真位移和实验位移的平方差。RADIOSS的求解结果表明,采用优化后的鸟体参数的鸟体模型的仿真曲线与实验曲线拟合效果大大提高。The accuracy of the bird model parameters has a signficant effect on bird strike simulation, and the inversion of physical parameters can overcome the limitations of artificial trial method,finding out reasonable bird model parameters and improving the accuracy of bird strike simulation. Based on the Hyper Study,a multidisciplinary optimization platform,orthogonal design of experimental（ DOE ）method is introduced to identify the model parameters which are sensitive to displacement, to simplify the complexity of optimization and to reduce the computation.Then those parameters are optimized based on adaptive response surface method （ARSM） ,and the optimization objective is to minimize the squared differences of simulation displacement and experimental displacement on the bird strike points.The results solved by RADIOSS demonstrate that when bird model adoptes the optimized parameters ,the simulation curve and experimental curve fit better.

关 键 词：参数反演 实验设计方法 自适应响应面法 HyperStudy 

分 类 号：TH16[机械工程—机械制造及自动化]