Desarrollo de la meta-heurística VMO en la plataforma de optimización Opt4J
Cargando...
Fecha
Autores
Almeida Montesino, Adonnis
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas
Resumen
Las técnicas meta-heurísticas han demostrado ser una herramienta completa para resolver problemas de optimización complejos; que aportan un equilibrio de “ buenas “ soluciones (relativamente cerca del óptimo global) y el tiempo y costo asequible. Unido a que como solución, el paradigma orientado a objetos se ha convertido en un mecanismo exitoso utilizado para aliviar la carga del desarrollo de la aplicación y, en particular, en la adaptación de una determinada meta-heurísticas al problema específico a resolver. Los marcos de trabajo para la optimización de meta-heurísticas son uno de los enfoques en ese sentido. El grupo de investigación de Inteligencia Artificial del CEI ha desarrollado varias heurísticas y meta-heurísticas de forma independiente. Resulta entonces conveniente el desarrollo de una metodología que haciendo uso de un marco de trabajo de optimización, permita el rápido desarrollo de nuevas meta-heurísticas así como su evaluación. En este trabajo se ha seleccionado el marco de trabajo Opt4J al que se le incorporará la meta-heurísticas VMO desarrollada por el grupo de IA.
The meta-heuristic techniques have proven to be a complete tool for solving complex optimization problems; they provide a balance of "good" solutions (relatively close to the global optimum) and time and affordable. Considering also that as a solution, the object-oriented paradigm has become a successful mechanism used to ease the burden of application development and, in particular, in the adaptation of a certain meta-heuristics to solve specific problems. Frameworks for optimization of meta-heuristics are one approach in this regard. The Artificial Intelligence research group belonging to the Studies Center of Informatics has developed several heuristics and meta-heuristics independently. It is therefore desirable to develop a methodology using a framework of optimization, enabling the rapid development of new meta-heuristics and evaluation. In this paper was selected the Opt4J framework that will incorporate the meta-heuristics VMO developed by the group of IA.
The meta-heuristic techniques have proven to be a complete tool for solving complex optimization problems; they provide a balance of "good" solutions (relatively close to the global optimum) and time and affordable. Considering also that as a solution, the object-oriented paradigm has become a successful mechanism used to ease the burden of application development and, in particular, in the adaptation of a certain meta-heuristics to solve specific problems. Frameworks for optimization of meta-heuristics are one approach in this regard. The Artificial Intelligence research group belonging to the Studies Center of Informatics has developed several heuristics and meta-heuristics independently. It is therefore desirable to develop a methodology using a framework of optimization, enabling the rapid development of new meta-heuristics and evaluation. In this paper was selected the Opt4J framework that will incorporate the meta-heuristics VMO developed by the group of IA.