Информационная система "Конференции"



International Conference on Numerical Methematics ICCM-2002


Abstracts


Numerical solution of differential and integral equations

Optimal on accuracy methods for approximate solving of the second kind Volterra integral equations for multiprocessor computers

Boykov I.V., Tynda A.N.

Penza State University (Penza)

Since many new applications of the Volterra integral equations have appeared for the last years, numerical methods for these equations are actively developing. This paper is dedicated to construction the optimal on the accuracy order methods of approximate solving the multidimensional Volterra and Abel-Volterra integral equations, based on the polynomial splines approximation of the exact solution. We consider the equations of the following types $$ x(t_1,ldots,t_l)+$$ $$ +intlimits_{0}^{t_l}cdotsintlimits_{0}^{t_1} h(t_1,ldots,t_l,tau_1,ldots,tau_l)g(t_1-tau_1,ldots,t_l-tau_l) x(tau_1,ldots,tau_l)dtau_1cdots dtau_l= $$ $$=f(t_1,ldots,t_l),$$ where $0le t_1,ldots,t_lle T,; h(t_1,ldots,t_l,tau_1,ldots,tau_l)$ and $f(t_1,ldots,t_l)$ are the functions, having partial derivatives till certain order $m$. Weakly singular kernels $g(t_1-tau_1,ldots,t_l-tau_l)$ may have the form $$ g(t_1,ldots,t_l)=t_1^{r_1+alpha_1}cdots t_l^{r_l+alpha_l}, ; 0 < alpha_i < 1 $$ or $$ g(t_1,ldots,t_l)=(t_1^2+cdots+t_l^2)^{r+alpha}, ; 0 < alpha < 1, $$ or $g(t_1,ldots,t_l)equiv 1$ in the regular equation case. Here we present the algorithm of parallelization of the constructed schemes to realize them on the computers with two and more processors. vspace{1cm} The work has been supported by Russian Humanitarian Scientific Fund (RHSF) (Research Grant Nr. 01-02-00147a)

Note. Abstracts are published in author's edition


Mail to Webmaster
|Home Page| |English Part| [SBRAS]
Go to Home
© 1996-2000, Siberian Branch of Russian Academy of Sciences, Novosibirsk
    Last update: 06-Jul-2012 (11:45:20)