Numerical methods improvements and large-scale computing techniques for electromagnetic simulations

来源 :第八届工业与应用数学国际大会 | 被引量 : 0次 | 上传用户:truebyb
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  In order to simulate complicated EM phenomena,such as inhomogeneous medium,curve interface and boundary,complicated BCs,many new techniques need to be developed and old ones be extended,such as DG,Matched Interface and Boundary(MIB)method,high-order algorithms,Method of Moment(MOM),Fast Algorithms,etc.
其他文献
Diesel engine performance and emissions are strongly coupled with fuel atomization and spray processes,which in turn are strongly influenced by injector flow dynamics.Modern engines employ micro-orifi
The development of new fuels from biomass is one promising technology for future energy supply.In the design process,the mixing efficiency of such novel biofuels needs to be characterized.The analysis
This paper presents results from an experimental investigation of the macroscopic and microscopic atomization and combustion behavior of B99 biodiesel,ethanol,and B99-ethanol blended fuels.In addition
We have experimentally found the liquid film flow characteristics rising along the outer surface of a rotating cone.When the film flow goes up fully upward,the liquid film cannot keep the filmwise con
We study multiple algorithms to select regularization parameters to stabilize deconvolution in high-speed communication measurement applications,investigating these algorithms in the presence of unspe
会议
We present some recent development of reduced-order modeling techniques in the fields of uncertainty quantification and computational fluid dynamics.We consider multilevel and weighted algorithms in t
会议
We focus on the following questions of random graph and complex networks.How to classify the structure of different random growing networks? How do the dynamical processes taking place on a random net
会议
Algorithmic scalability to high dimensional parameters and computationalefficiency of numerical solvers are two significant challenges in largescale,PDE-constrained inverse problems.
会议
The generalized Polynomial Chaos(gPC)method is a popular method for solving partial differential equations(PDEs)with random parameters.However,when the probability space has high dimensionality,the so
会议
The Laplacian eigenvalue of a graph plays an important role in controlling other graph parameters.It is closed related to random walks.In this talk,we introduce a set of Laplacian eigenvalues of a uni
会议