期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
A POSTERIORI ERROR ANALYSIS OF A FULLY-MIXED FINITE ELEMENT METHOD FOR A TWO-DIMENSIONAL FLUID-SOLID INTERACTION PROBLEM
1
作者 Carolina Dominguez Gabriel N. Gatica salim meddahi 《Journal of Computational Mathematics》 SCIE CSCD 2015年第6期606-641,共36页
In this paper we develop an a posteriori error analysis of a fully-mixed finite element method for a fluid-solid interaction problem in 2D. The media are governed by the elas-todynamic and acoustic equations in time-... In this paper we develop an a posteriori error analysis of a fully-mixed finite element method for a fluid-solid interaction problem in 2D. The media are governed by the elas-todynamic and acoustic equations in time-harmonic regime, respectively, the transmission conditions are given by the equilibrium of forces and the equality of the corresponding normal displacements, and the fluid is supposed to occupy an annular region surrounding the solid, so that a Robin boundary condition imitating the behavior of the sommerfeld condition is imposed on its exterior boundary. Dual-mixed approaches are applied in both domains, and the governing equations are employed to eliminate the displacenent a of the solid and the pressure p of the fluid. In addition, since both transmission conditions become essential, they are enforced weakly by means of two suitable lahrange multipliers The unknowns of the solid and the fluid are then approximated by a conforming galerkin scheme defined in terms of PEERS elements in the solid, Raviart-Thomas of lowest order in the fluid, and continuous piecewise linear functions on the boundary, As the main con-tribution of this work, we derive a reliable and efficient residual-based a posteriori error estimator for the aforedescribed coupled problem. Some numerical results confirming the properties of the estimator are also reported. 展开更多
关键词 Mixed finite elements Helmholtz equation Elastodynamic equation A poste- riori error analysis
原文传递
基于声音多特征贝叶斯网络融合的话者识别研究 被引量:14
2
作者 朱坚民 张雷 +1 位作者 翟东婷 雷静桃 《仪器仪表学报》 EI CAS CSCD 北大核心 2013年第9期2058-2067,共10页
针对基于语音单一特征提取方法所存在的话者识别准确率较低的问题,提出将话者语音中反映人耳听觉感知特性的MFCC特征和接近心理声学临界频带的1/3倍频程(1/3 octave)特征作为话者声音的特征参数,设计话者识别的贝叶斯网络,融合2种声音... 针对基于语音单一特征提取方法所存在的话者识别准确率较低的问题,提出将话者语音中反映人耳听觉感知特性的MFCC特征和接近心理声学临界频带的1/3倍频程(1/3 octave)特征作为话者声音的特征参数,设计话者识别的贝叶斯网络,融合2种声音特征参数,通过贝叶斯网络推理实现话者识别。贝叶斯网络通过学习过程确定已注册话者各声音特征的条件概率。进行话者识别时,贝叶斯网络利用贝叶斯定理及条件独立性假设融合待识别话者声音的MFCC特征和1/3倍频程特征,计算每个已注册话者对输入语音特征矢量的后验概率,根据后验概率的大小实现待识别话者的推断。话者识别实验结果表明:提出的基于声音多特征贝叶斯网络融合的话者识别方法可行有效,识别正确率达到100%。 展开更多
关键词 MFCC特征 1 3倍频程特征 贝叶斯网络 后验概率 话者识别 融合
在线阅读 下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部