首页 | 学院概况 | 师资队伍 | 本科生教育 | 研究生教育 | 科学研究 | 交流合作 | 党建工作 | 学生天地 | 工会视窗 | 校友之声 | 院务公开 
首页
 学院动态 
 学术交流 
 教育教学 
 学生工作 
 热点聚焦 
首页>>学术交流>>正文
航空学院青年学术论坛第六十三讲(两次讲座)
2017-05-14 06:59  

报告人:Dr Sergei Kucherenko

讲座一

题目:基于偏导的测度指标和随机搜索优化算法

时间和地点: 2017.5.15下午2:30-4:00 航空楼A706

摘要: 基于方差的全局灵敏度分析能够高效地分析输入变量不确定性对输出响应的全面影响。MC和QMC可求解收敛的测度指标。基于偏导的测度指标能够有效地界定全局灵敏度指标的上限,并给出与全局灵敏度指标相同的重要性分析。在此基础上,建立更为有效的上下限分析,以期确定更为严谨的总指标取值范围。

HDMR 为直接提供测度指标提供了便捷方式。通过积分和回归方法建立HDMR的求解过程。改进的GMDH方法耦合HDMR可进一步高效求解全局灵敏度指标。

随机截断优化算法可以有效搜索所有全局最优解和尽可能多的局部最优解,并可用于含连续和离散变量的问题,为工程设计提供更多的可行性方案。

讲座二

题目:Global sensitivity analysis, metamodeling and optimization algorithm

时间:5月16日上午9:30-11:00

地点:航空楼A706

摘要:

Sensitivity can reflect the effect of input variables on output response. Global sensitivity analysis (GSA) is a very useful tool to evaluate the influence of input variables in the whole distribution range. Sobol’ method is the most commonly used among variance-based methods, which are efficient and popular GSA techniques. Many simulation methods can calculate the Sobol’ sensitivity indices, such as the Monte Carlo (MC) method or the quasi-MC (QMC) method using Sobol’ sequence. High dimensional model representation (HDMR) is a popular way to compute Sobol’ indices. Derivative-based global sensitivity measure (DGSM) is proposed and discovered the link between DGSM and the total sensitivity index.

In the cases of computationally expensive models, the metamodelling technique which maps inputs and outputs is a very useful and practical way of making computations tractable. A number of new techniques which improve the efficiency of the Random Sampling-High dimensional model representation (RS-HDMR) for models with independent and dependent input variables are presented. Two different metamodelling methods for models with dependent input variables are compared. Both techniques are based on Quasi Monte Carlo variant of RS-HDMR. The first technique makes use of transformation of the dependent input vector into a Gaussian independent random vector and then applying decomposition of the model using a tensored Hermite polynomial basis. The second approach uses a direct decomposition of the model function into a basis which consists of the marginal distributions of input components and their joint distribution.

报告人简介:Sergei Kucherenko, 生于1960年,自2000年至今就职于英国帝国学院(原帝国理工)研究员,兼任欧盟框架委员会成员。1984年获国家核科学大学(前莫斯科工程物理研究所)博士学位,导师是全局灵敏指标的提出者Sobol’ I M。研究方向:数学/数值计算、软件工程,主要研究内容包括:全局灵敏度测度分析、随机全局优化算法、代理模型等,在国际知名期刊(Computer Physics CommunicationsReliability Engineering and System Safety等)发表论文60多篇次,其中SCI 12区论文十余篇,被引用量超过千余次,编制多个共享软件,用户达3万余人,两次获得EPSRC (Engineering and Physical Sciences Research Council, 英国工程与自然科学研究理事会)的基金经费支持,组织并参与前八届International Conference on Sensitivity Analysis of Model Output会议,担任多个国际期刊的特约审稿人。https://www.researchgate.net/profile/Sergei_Kucherenko2

 

关闭窗口
版权所有©西北工业大学航空学院 | http://hangkong.nwpu.edu.cn