A simulated annealing algorithm for the robust decomposition of temporal horizons in production planning problems

The problem of robust decomposition of temporal horizons in production planning was first introduced by Torres [1]. Later, in [2], Torres suggests to start with an integer solution found by dynamic programming, and then to use a simulated annealing algorithm to improve it. According to [2], more need...

Full description

Autores:
Torres Delgado, José Fidel
Vélez Gallego, Mario César
Tipo de recurso:
Fecha de publicación:
2007
Institución:
Universidad EAFIT
Repositorio:
Repositorio EAFIT
Idioma:
spa
OAI Identifier:
oai:repository.eafit.edu.co:10784/14541
Acceso en línea:
http://hdl.handle.net/10784/14541
Palabra clave:
Production Planning
Simulated Annealing
Planificación De Producción
Recocido Simulado
Rights
License
Copyright (c) 2007 José Fidel Torres Delgado, Mario César Vélez Gallego
Description
Summary:The problem of robust decomposition of temporal horizons in production planning was first introduced by Torres [1]. Later, in [2], Torres suggests to start with an integer solution found by dynamic programming, and then to use a simulated annealing algorithm to improve it. According to [2], more needs to be known about the impact of the control parameters in the simulated annealing algorithm, and their sensitivity with respect to the quality of the solutions. In this work we develop this idea and analyze in depth the ability of the simulated annealing algorithm to improve the initial solution. As a result of the computational experiments conducted, we determined that the cooling scheme and the cooling rate have significant effect on the quality of the final solution. It was also established that the solution found depends strongly on the characteristics of the operations plan, finding better solutions for plans with shorter temporal horizons.