Selective methodology of population dynamics for optimizing a multiobjective environment of job shop production

This paper develops a methodology based on population genetics to improve the performance of two or more variables in job shop production systems. The methodology applies a genetic algorithm with special features in the individual selection when they pass from generation to generation. In comparison...

Full description

Main Authors: Ruiz, Santiago, Castrillón, Omar Danilo, Sarache, William Ariel
Format: Artículo
Language: Español
Published: 2015
Subjects:
Online Access: http://revistas.ucr.ac.cr/index.php/matematica/article/view/17558
http://hdl.handle.net/10669/13066
Summary: This paper develops a methodology based on population genetics to improve the performance of two or more variables in job shop production systems. The methodology applies a genetic algorithm with special features in the individual selection when they pass from generation to generation. In comparison with the FIFO method, the proposed methodology showed better results in the variables makespan, idle time and energy cost. When compared with NSGA II, the methodology did not showed relevant differences in makespan and idle time; however better performance was obtained in energy cost and, especially, in the number of required iterations to get the optimal makespan.