info:eu-repo/semantics/bookPart
A Hybrid Simulation Based on Multi-Objective Algorithm for Manufacturing Cells Optimization
Autor
Pedro Pérez Villanueva
Elias Gabriel Carrum Siller
Resumen
This paper presents a hybrid simulation based on multi-objective algorithm for creation and optimization of manufacturing cells; these cells are created by using the principle of group technology with binary matrices. The algorithm used in this paper is the NSGA-II using a seed made by a modified ART neural network, the NSGA-II algorithm is used to maximize the final inventory, minimize the WIP, and minimize the movement time in order to create an optimized cells, after that, the best solution is compared using simulation against the original matrices, the cell formation given by and modify ART neural network and the NSGA-II algorithm without the seed. The solution given by the hybrid NSGA-II algorithm gives superiors solutions when the seed is used.