A novel stochastic collocation method for uncertainty propagation in complex mechanical systems

IMR OpenIR

	A novel stochastic collocation method for uncertainty propagation in complex mechanical systems
其他题名	A novel stochastic collocation method for uncertainty propagation in complex mechanical systems
	Qi WuChao 1; Tian SuMei 1; Qiu ZhiPing 2
	2015
发表期刊	SCIENCE CHINA-PHYSICS MECHANICS & ASTRONOMY
ISSN	1674-7348
卷号	58 期号:2
摘要	This paper presents a novel stochastic collocation method based on the equivalent weak form of multivariate function integral to quantify and manage uncertainties in complex mechanical systems. The proposed method, which combines the advantages of the response surface method and the traditional stochastic collocation method, only sets integral points at the guide lines of the response surface. The statistics, in an engineering problem with many uncertain parameters, are then transformed into a linear combination of simple functions' statistics. Furthermore, the issue of determining a simple method to solve the weight-factor sets is discussed in detail. The weight-factor sets of two commonly used probabilistic distribution types are given in table form. Studies on the computational accuracy and efforts show that a good balance in computer capacity is achieved at present. It should be noted that it's a non-gradient and non-intrusive algorithm with strong portability. For the sake of validating the procedure, three numerical examples concerning a mathematical function with analytical expression, structural design of a straight wing, and flutter analysis of a composite wing are used to show the effectiveness of the guided stochastic collocation method.
其他摘要	This paper presents a novel stochastic collocation method based on the equivalent weak form of multivariate function integral to quantify and manage uncertainties in complex mechanical systems. The proposed method, which combines the advantages of the response surface method and the traditional stochastic collocation method, only sets integral points at the guide lines of the response surface. The statistics, in an engineering problem with many uncertain parameters, are then transformed into a linear combination of simple functions’ statistics. Furthermore, the issue of determining a simple method to solve the weight-factor sets is discussed in detail. The weight-factor sets of two commonly used probabilistic distribution types are given in table form. Studies on the computational accuracy and efforts show that a good balance in computer capacity is achieved at present. It should be noted that it’s a non-gradient and non-intrusive algorithm with strong portability. For the sake of validating the procedure, three numerical examples concerning a mathematical function with analytical expression, structural design of a straight wing, and flutter analysis of a composite wing are used to show the effectiveness of the guided stochastic collocation method.
关键词	RELIABILITY-ANALYSIS POLYNOMIAL CHAOS DIFFERENTIAL-EQUATIONS DECOMPOSITION CONVERGENCE SPEED uncertainty propagation stochastic collocation method equivalent weak form guided surface statistics
收录类别	CSCD
语种	英语
资助项目	[Defense Industrial Technology Development Program] ; [National Natural Science Foundation of China]
CSCD记录号	CSCD:5386338
引用统计
文献类型	期刊论文
条目标识符	http://ir.imr.ac.cn/handle/321006/150120
专题	中国科学院金属研究所
作者单位	1.中国科学院金属研究所 2.Beijing Univ Aeronaut & Astronaut, Institute Solid Mech, Beijing 100191, Peoples R China
推荐引用方式 GB/T 7714	Qi WuChao,Tian SuMei,Qiu ZhiPing. A novel stochastic collocation method for uncertainty propagation in complex mechanical systems[J]. SCIENCE CHINA-PHYSICS MECHANICS & ASTRONOMY,2015,58(2).
APA	Qi WuChao,Tian SuMei,&Qiu ZhiPing.(2015).A novel stochastic collocation method for uncertainty propagation in complex mechanical systems.SCIENCE CHINA-PHYSICS MECHANICS & ASTRONOMY,58(2).
MLA	Qi WuChao,et al."A novel stochastic collocation method for uncertainty propagation in complex mechanical systems".SCIENCE CHINA-PHYSICS MECHANICS & ASTRONOMY 58.2(2015).