|
|
Prostory funkcí, diferenciální operátory a nelineární analýza
Drábek, P. ; Rákosník, Jiří
The proceedings contain four of the invited main lectures and 23 selected contributions by the participants. All published papers were subject to the standard referee procedure. The conference was the sixth in the series of international meetings organized under the same name alternatively in Finland, Germany and Czech Republic since 1988. This conference was devoted to the seventieth birthday of Prof. Alois Kufner, the founder of the Czech school on the theory of function spaces.
|
|
|
Numerická kvadratura vyššího řádu ve 2D a 3D
Segeth, Karel ; Šolín, P. ; Doležel, I.
The construction of higher-order Gauss quadrature rules in 2D and 3D is it considered. Many open questions are related to the desing of (symmetric) formulae with minimum number of points. After outlining some of the state-of-the- art knowledge and open problems in contemporary numerical quadrature related to finite element methods, we present and discuss several techniques which can (and should) be used where the current knowledge of Gaussian quadrature rules is not sufficient.
|
|
|
Numerická kvadratura pro metody konečných prvků vyšších prvků
Šolín, P. ; Segeth, Karel ; Doležel, I.
The importance of suitable numerical quadrature is usually not emphasized in the context of higher-order finite element methods or their p- or hp-adaptive versions. However, quadrature lies at the heart of finite element codes and it influences their performance in a crucial way. After outlining some of the state-of-the-art knowledge and open problems in contemporary numerical quadrature related to finite element methods, we present and discuss several techniques which can (and should) be used where the current knowledge of Gaussian quadrature rules is not sufficient
|
|
|
Tři způsoby interpolace na konečných prvcích
Šolín, Pavel ; Segeth, Karel
Interpolation on finite elements usually occurs in a Hilbert space setting, which means that interpolation techniques involving orthogonal projection are an alternative for the traditional Lagrange nodal interpolation schemes. In addition to the Lagrange interpolation, this paper discusses the global orthogonal projection and the projection-based interpolation. These techniques are compared from the point of view of quality, efficiency, sensitivity to input parameters and other aspects.
|
|
|
Programy a algoritmy numerické matematiky 12
This book contains more than 30 papers presented at the seminar Programs and Algorithms of Numerical Mathematics held in Dolní Maxov, Czech Republic, June 6-11, 2004. The contributions deal mostly with the finite element method and its applications. Other subjects, as spline construction, numerical linear algebra algorithms, or optimization, for example, are also covered.
|
|
| |
|
|
Rychlý a zaručený aposteriorní odhad chyby
Vejchodský, Tomáš
The equilibrated residual method and the method of hypercircle are popular methods for a posteriori error estimation for linear elliptic problems. Both these methods are intended to produce guaranteed upper bounds of the energy norm of the error, but the equilibrated residual method is guaranteed only theoretically. The disadvantage of the hypercircle method is its globality, hence slowness. The combination of these two methods leads to local, hence fast, and guaranteed a posteriori error estimator.
|
|
|
Extrapolace. Aktuální výsledky a problémy
Krbec, Miroslav
This paper is deals with several recent extrapolation results from the recent author´s papers and tackles further the relations of the spaces involved. First, very simple and transparent proofs of extrapolation theorems of Yano type in Lorentz spaces is given, unifiying classical and new results. Second, the variation of the classical situation near arbitrary L_p is studied and the answer is icen in terms of Zygmund, Lorentz-Zygmund and small Lebesgue spaces.
|
|
| |
|
| |
guest ::
