The code files used to get the results published in the paper "Numerical reliability in the simulation of Chaotic Systems using Matlab and C ++" (in portuguese).
Download the manuscript here.
This work compares the numerical reliability in simulating chaotic systems using C++ and Matlab. The Lower Bound Error was used as a parameter to measure how reliable the simulations are. The chaotic systems of Lorenz and logistic map were used as case studies. With the results one could note that, as expected, C++ is faster than Matlab. However, this paper's main contribution is that, for the same precision (64 bits), C++ shows a more reliable result in terms of the quality of the simulation in such a way that, accordingly to the proposed metric (simulation time), resulted in a gain up to 26% for the Lorenz system and up to 54% for the Logistic map, when compared to Matlab results.
Authors: Wilson Rocha Lacerda Junior, Maurício Silva Barbosa, Samir Angelo Milani Martins e Erivelton Geraldo Nepomuceno.
Website: http://www.ufsj.edu.br/gcom
Please send suggestions for improvement of the above code to Wilson Rocha at this email address: wilsonrljr@outlook.com
@article{junior2017confiabilidade,
title={Confiabilidade num{\'e}rica na simula{\c{c}}ao de Sistemas Ca{\'o}ticos utilizando Matlab e C+},
author={Junior, Wilson Rocha Lacerda and Barbosa, Maur{\i}cio Silva and Nepomuceno, Erivelton Geraldo and Martins, Samir Angelo Milani},
journal={Revista de Engenharias da Faculdade Salesiana},
number={6},
pages={2--9},
year={2017}
}