Shear strength prediction of reinforced-concrete beams based on fuzzy theory

An alternative analytical method based on fuzzy theory was developed for the accurate evaluation of the shear strength of reinforced-concrete (RC) beams. In total, 1014 experimental results of shear tests, covering a wide range of design parameters of simply supported RC slender and deep beams, were used for training and validation of the proposed fuzzy-based model. Bayesian analysis was performed to determine the input parameters used in the fuzzy model and a data distribution index was developed to evaluate the evenness of the training data distribution. A model based on fuzzy rules and using fuzzy membership functions was developed and the strengths predicted by the proposed model were compared with those predicted by current design codes (ACI 318-11 and Eurocode 2). The results showed that fuzzy set theory can properly address the complex interaction between various modelling parameters and the fuzzy rule based model enhanced the prediction of shear strength.