


Paper's Title:
Accuracy of Implicit DIMSIMs with Extrapolation
Author(s):
A. J. Kadhim, A. Gorgey, N. Arbin and N. Razali
Mathematics Department, Faculty of
Science and Mathematics,
Sultan Idris Education University,
35900 Tanjong Malim, Perak,
Malaysia
Email: annie_gorgey@fsmt.upsi.edu.my
Unit of Fundamental Engineering Studies,
Faculty of Engineering, Built Environment,
The National University of Malaysia,
43600 Bangi,
Malaysia.
Abstract:
The main aim of this article is to present recent results concerning diagonal implicit multistage integration methods (DIMSIMs) with extrapolation in solving stiff problems. Implicit methods with extrapolation have been proven to be very useful in solving problems with stiff components. There are many articles written on extrapolation of RungeKutta methods however fewer articles on extrapolation were written for general linear methods. Passive extrapolation is more stable than active extrapolation as proven in many literature when solving stiff problems by the RungeKutta methods. This article takes the first step by investigating the performance of passive extrapolation for DIMSIMs type2 methods. In the variable stepsize and order codes, order2 and order3 DIMSIMs with extrapolation are investigated for Van der Pol and HIRES problems. Comparisons are made with ode23 solver and the numerical experiments showed that implicit DIMSIMs with extrapolation has greater accuracy than the method itself without extrapolation and ode23.
Paper's Title:
An Approximation of Jordan Decomposable Functions for a Lipschitz Function
Author(s):
Ibraheem Alolyan
Mathematics Department,
College of Science, King Saud University
P.O.Box: 2455, Riyadh 11451,
Saudi Arabia
ialolyan05@yahoo.com
Abstract:
The well known Jordan decomposition theorem gives the useful characterization that any function of bounded variation can be written as the difference of two increasing functions. Functions which can be expressed in this way can be used to formulate an exclusion test for the recent Cellular Exclusion Algorithms for numerically computing all zero points or the global minima of functions in a given cellular domain [2,8,9]. In this paper we give an algorithm to approximate such increasing functions when only the values of the function of bounded variation can be computed. For this purpose, we are led to introduce the idea of εincreasing functions. It is shown that for any Lipschitz continuous function, we can find two εincreasing functions such that the Lipschitz function can be written as the difference of these functions.
Search and serve lasted 1 second(s).