Elaboración de un modelo evolutivo híbrido de algoritmos genéticos y redes neuronales artificiales para dosificaciones de mezclas de concreto reforzadas con fibras metálicas

Abstract

An evolutionary model is developed in a computing environment to propose metal fiber reinforced concrete mixture dosages for compressive strength applications. The model is hybrid as it includes both a dosage system based on genetic algorithm and a properties prediction system based on artificial neural networks. The results obtained are compared with experimentally reported dosages set, and the comparisons show an approximation in the simulation process. Given the characteristics of the model, it is considered a contribution to concrete technology. (English) [ABSTRACT FROM AUTHOR]