Application of group method of data handling to Estimate the Shear Strength of RC Beams Reinforced with FRP Bars

In spite of the high abilities of neural networks in solving complex problems and their increasing place in engineering issues, the determination of a simple mathematical structure for the application of their final models is difficult. In contrast, the GMDH method, which is based on second-order polynomials, can be chosen in many cases as an alternative and appropriate solution. In this paper, the ability of GMDH to evaluate the shear strength of concrete beams with FRP materials has been evaluated. For this purpose, some inputs are used, and the goal is prediction shear strength. A set of experimental results has been used to determine the polynomial coefficients. The final model has two middle layers, which consists of three and two polynomials in the first and second layers. This model is validated using laboratory data and its results are presented at the end. In summary, the proposed model in this paper has a suitable performance.

Keywords:

FRP , GMDH , RC beam , Shear strength.