Desarrollo de un algoritmo memético basado en VMO
Archivos
Fecha
2010-07-14
Autores
Martínez Pérez, Carlos Alberto
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas
Resumen
Los métodos aproximativos para la resolución de problemas de optimización se han venido desarrollando debido alta complejidad que dichos problemas han alcanzado. Una buena explotación de las soluciones ya encontradas, así como una correcta exploración del espacio de búsqueda son factores imprescindibles para obtener soluciones de calidad. Esta investigación está encaminada a desarrollar un modelo que combine de manera equilibrada estos factores, tomando como base poblacional para la exploración el modelo basado en Mallas Variables e incorporándole al mismo una etapa de Búsqueda Local encargada de explotar las regiones cercanas a las buenas soluciones ya encontradas. Con el objetivo de presentar un modelo más robusto que el modelo basado en Mallas Variables puro, que obtenga soluciones de mejor calidad que el mismo. Los resultados de esta investigación son la obtención de distintas variantes de un Algoritmo Memético con mejor comportamiento que el modelo basado en Mallas Variables puro para la resolución de problemas continuos.
The approximate methods for the resolution of problems of optimization one has come developing due to the high complexity that these problems have reached. A good exploitation of the good solutions and a correct exploration of the search space are indispensables factors to obtain solutions of quality. This investigation is guided to develop a model that combines in a balanced way these factors, taking like populational base, for the exploration the model based on Variable Meshes and incorporating to the same one a stage of Local Search in charge of exploiting the near regions to the good ones already solves. The aim of this research is presenting a more robust model that the based on Variable Meshes pure, that it obtains solutions of better quality that the pure model. The results of this investigation are the obtaining of different variants of a Memetic Algorithm with better behavior that the model based on Variable Meshes pure for the resolution of continuous problems.
The approximate methods for the resolution of problems of optimization one has come developing due to the high complexity that these problems have reached. A good exploitation of the good solutions and a correct exploration of the search space are indispensables factors to obtain solutions of quality. This investigation is guided to develop a model that combines in a balanced way these factors, taking like populational base, for the exploration the model based on Variable Meshes and incorporating to the same one a stage of Local Search in charge of exploiting the near regions to the good ones already solves. The aim of this research is presenting a more robust model that the based on Variable Meshes pure, that it obtains solutions of better quality that the pure model. The results of this investigation are the obtaining of different variants of a Memetic Algorithm with better behavior that the model based on Variable Meshes pure for the resolution of continuous problems.
Descripción
Palabras clave
Optimización, Modelo de Mallas Variables, Búsqueda Local, Hibridación, Algoritmo Memético, Inteligencia Artificial