APLICAÇÃO DE ALGORITMOS DE APRENDIZADO DE MÁQUINA NA ESTIMATIVA DO FATOR DE SEGURANÇA DE TALUDES COM DADOS SIMULADOS PARA DIFERENTES MÉTODOS DE EQUILÍBRIO LIMITE

Abstract

Slope stability is a fundamental aspect of geotechnical engineering and plays a crucial role in ensuring the safety and performance of civil and environmental engineering projects. Traditional slope stability assessments rely on limit equilibrium methods, such as those developed by Fellenius, Bishop, Janbu, and others, which evaluate the safety factor based on simplified assumptions about soil behavior. With the evolution of computational tools, geotechnical software has made it possible to conduct more realistic and complex simulations, enhancing the reliability of engineering designs. Recently, artificial intelligence techniques, especially machine learning, have emerged as promising tools for improving slope stability analysis by better addressing uncertainties and reducing the limitations of conventional methods. This study proposes the development of a predictive model using machine learning algorithms trained on simulated data generated from geotechnical software. The objective is to evaluate the performance of different algorithms in estimating the safety factor of slopes and to compare their accuracy and efficiency with traditional methods. The research aims to contribute to the advancement of AI applications in geotechnical engineering and to promote safer, more efficient design practices.