|
|
Parallel Computer Systems Based on Numerical Integrations
Kraus, Michal ; Kubátová, Hana (referee) ; Kollár,, Ján (referee) ; Kunovský, Jiří (advisor)
This thesis deals with continuous system simulation. The systems can be described by system of differential equations or block diagram. Differential equations are usually solved by numerical methods that are integrated into simulation software such as Matlab, Maple or TKSL. Taylor series method has been used for numerical solutions of differential equations. The presented method has been proved to be both very accurate and fast and also procesed in parallel systems. The aim of the thesis is to design, implement and compare a few versions of the parallel system.
|
|
|
Methods of numerical integration
Čoupek, Filip ; Tomášek, Petr (referee) ; Zatočilová, Jitka (advisor)
This bachelor thesis focuses on numerical calculation of a simple specific integral. First, the basic concepts are established and briefly described interpolation and orthogonal polynomials, from which the individual formulas are based. Emphasis is placed on the introduction, derivation and description of Newton-Cotes quadrature formulas, Gauss quadrature formulas and Clenshaw-Curtis quadrature formulas. In the penultimate chapter we describe the principle of adaptive integration and Romberg's method. At the end of the thesis is a comparison of individual methods on specific examples using the software Matlab.
|
|
|
Partial Differential Equations Parallel Solutions
Čambor, Michal ; Kunovský, Jiří (referee) ; Šátek, Václav (advisor)
This thesis deals with the concepts of numerical integrator using floating point arithmetic for solving partial differential equations. The integrator uses Euler method and Taylor series. Thesis shows parallel and serial approach to computing with exponents and significands. There is also a comparison between modern parallel systems and the proposed concepts.
|
|
|
Methods of numerical integration
Čoupek, Filip ; Tomášek, Petr (referee) ; Zatočilová, Jitka (advisor)
This bachelor thesis focuses on numerical calculation of a simple specific integral. First, the basic concepts are established and briefly described interpolation and orthogonal polynomials, from which the individual formulas are based. Emphasis is placed on the introduction, derivation and description of Newton-Cotes quadrature formulas, Gauss quadrature formulas and Clenshaw-Curtis quadrature formulas. In the penultimate chapter we describe the principle of adaptive integration and Romberg's method. At the end of the thesis is a comparison of individual methods on specific examples using the software Matlab.
|
|
|
Multiple Integrals
Valešová, Nikola ; Veigend, Petr (referee) ; Šátek, Václav (advisor)
The problem of definite integral and differential equation computation is still a significant part of many scientific branches and the solution of integral calculus tasks can be found in many industrial fields too. During the computation of such tasks, the accuracy and high-speed requirements are often confronted. These requirements are crucial during the process of the suitable method choice. The aim of this thesis is to propose, describe, implement and test a new numerical method, which combines the solution of definite integrals by transforming them into differential equations solved by the Taylor series with the traditional methods, which use the Newton-Cotes formulas. As a result, a new application has been developed, that provides fast results of definite two-dimensional integrals and reaches at least the precision of MATLAB. The major accomplishment of this thesis is the development of a new numerical method and its comparison to other established ways of computation.
|
|
|
Newton and numerical mathematics
Obrátil, Štěpán ; Nechvátal, Luděk (referee) ; Zatočilová, Jitka (advisor)
Topic of this bachelor thesis are Newton's methods for numerical solutions of various problems. Especially the problems of solving nonlinear equations and systems of nonlinear equations, as well as numerical integration are explained. The Newton's method for solving nonlinear equations is presented, as well as its many modifications and its generalisation for systems of nonlinear equations. Usefulness of methods is demonstrated on various examples. In the end, Newton-Cotes quadrature formulae for numerical integration are presented.
|
|
|
Parallel Computer Systems Based on Numerical Integrations
Kraus, Michal ; Kubátová, Hana (referee) ; Kollár,, Ján (referee) ; Kunovský, Jiří (advisor)
This thesis deals with continuous system simulation. The systems can be described by system of differential equations or block diagram. Differential equations are usually solved by numerical methods that are integrated into simulation software such as Matlab, Maple or TKSL. Taylor series method has been used for numerical solutions of differential equations. The presented method has been proved to be both very accurate and fast and also procesed in parallel systems. The aim of the thesis is to design, implement and compare a few versions of the parallel system.
|
|
|
Integration of microstructure informed enrichment base functions
Ladecký, Martin ; Zeman,, Jan (referee) ; Eliáš, Jan (advisor)
This thesis deals with problems related to the numerical integration of rapidly oscillatory functions. Analyze classical methods of numerical integration and compare them with method published by David Levin\cite{levin82}. Levin's method is applied in solving Laplace differential equation that describes deflection of membrane. To solve potential problem is used hybrid finite element method with Trefftz bases functions.
|
|
|
Physics Simulator for SandBox Type Games
Kotulič, Patrik ; Peringer, Petr (referee) ; Hrubý, Martin (advisor)
This thesis focuses on the creation of a physics simulator of rigid body mechanics. The reader is familiarized with the principles used in current simulators of rigid body mechanics. The thesis then deals with numerical integration of the equations of motion, collision detection and collision resolution. The simulator is presented on a series of experiments. The resulting application can be used as a video game or as a tool for experimentation.
|
|
|
Specialized Parallel System and Elementar Processors
Nevřala, Pavel ; Kunovský, Jiří (referee) ; Šátek, Václav (advisor)
The present thesis deals with the conceptual design of the elementary processor for computation of fixed point. This study will examine and analyze the cooperation of processors in the calculation of linear systems of differential equations with a focus on Euler's method, Runge-Kutta method and Taylor series. The thesis is divided into two major parts. The first part contains the processor design according to the type of communication processor internal components and the type of communication between processors in the calculation. The second part describes the design and implementation of a specialized multi-variant parallel system and a compararison between them.
|
guest ::
