FAI: Fast, accurate, and intelligent approach and prediction tool for flexural capacity of FRP-RC beams based on super-learner machine learning model

Abstract

Fiber-reinforced polymer (FRP) composites have recently been considered in the field of structural engineering as one of the best alternatives to conventional steel reinforcement due to their high tensile strength, lightweight, cost-effectiveness, and superior corrosion resistance. However, the variation in FRP physical and mechanical characteristics among the different FRP types and manufacturers makes it difficult to predict the strength of FRP-reinforced concrete (RC) members. For that reason, an efficient prediction tool was developed for a fast, accurate, and intelligent (FAI) prediction of the flexural capacity of FRP-RC beams based on the result of an optimized super-learner machine learning (ML) model. A database of the experimental results on the flexural strength of FRP-RC beams was compiled and randomly split into 80% train and 20% test sets. Six factors were considered in the model; namely, width and effective depth of the beam, concrete compressive strength, FRP flexural reinforcement ratio, FRP modulus of elasticity, and FRP ultimate tensile strength. Grid search is combined with a 10-fold cross-validation to optimize the hyperparameters of the ML models. The prediction capability of the proposed super-learner ML model was benchmarked against boosting- and tree-based ML models, such as classification and regression trees, adaptive boosting, gradient boosted decision trees, and extreme gradient boosting. Moreover, a comparison with the existing code and guideline equations showed that the proposed super-learner ML model provided the most desirable prediction of the flexural capacity of FRP-RC beams.