Generalized Simulated Annealing Algorithm for Matlab

Many problems in biology, physics, mathematics, and engineering, demand the determination of the global optimum of multidimensional functions. Simulated annealing is a meta-heuristic method that solves global optimization problems. There are three types of simulated annealing: i) classical simulated annealing; ii) fast simulated annealing and iii) generalized simulated annealing. Among them, generalized simulated annealing is the most efficient. Matlab is one of the most widely software used in numeric simulation and scientific computation. Matlab optimization toolbox provides a variety of functions able to solve many complex problems. In this article, the generalized simulated annealing method was described, the GSA function that contains this method was applied to some mathematical problems were solved in order to evaluate the efficiency of GSA with respect to some of Matlab optimization functions. As a result, it was found that the GSA function not only manages to be effective in its convergence to the global optimum but also it does so quickly. Likewise, it was observed that, in general terms, GSA was more efficient than the functions with which it was compared. Therefore, it can be concluded that the GSA function is a novel and effective alternative for addressing optimization problems using Matlab.

• Autores: Wilches Visbal, Jorge Homero; Martins Da Costa, Alessandro
• Fecha de publicación: 2019
• Institución: Universidad EAFIT
• Repositorio: Repositorio EAFIT
• Idioma: spa
• OAI Identifier: oai:repository.eafit.edu.co:10784/17659
• Acceso en línea: http://hdl.handle.net/10784/17659
• Palabra clave: Simulated annealing; Efficiency; Optimization; GSA; Matlab; Recocido simulado; Optimización; Eficiencia; GSA; Matlab
• License: Copyright © 2019 Jorge Homero Wilches Visbal, Alessandro Martins Da Costa