An approach involving the coupling of artificial neural network with the genetic algorithm optimizer for predicting the shear capacity of <scp>SFRC</scp> beams

Abstract While various code provision equations exist for evaluating the shear strength of steel fiber‐reinforced concrete (SFRC) beams, engineers involved in structural retrofitting and analysis often face challenges in identifying a practical yet accurate equation. This paper introduces a novel hybrid machine learning model known as Genetic Algorithm Optimizer‐artificial neural network (GAO‐ANN), integrating the capabilities of a multilayer perceptron ANN with the GAO. For this study, an extensive experimental database of slender SFRC beams was meticulously compiled. Utilizing this database and the proposed hybrid approach, a concise yet precise closed‐form equation was developed to assess the shear capacity of SFRC beams. Additionally, a set of existing equations was provided for comparison in terms of accuracy and capabilities. While the Sarveghadi et al. equation demonstrated superior accuracy compared to other existing equations, the GAO‐ANN equation emerged as a simpler yet equally accurate alternative. In predicting the shear capacity of SFRC beams during the design and planning stages of civil engineering projects, the hybrid approach known as GAO‐ANN proves to be highly effective. Beyond contributing to the advancement of structural engineering practices, this model provides engineers with a reliable and efficient method for evaluating shear capacity.