Modeling Torsional Strength of Reinforced Concrete Beams using Genetic Programming Polynomials with Building Codes

The potential of using genetic programming to predict engineering data has caught the attention of researchers in recent years. This paper utilizes a derivative of genetic programming to model the torsional strength of reinforced concrete beams using polynomial-like equations. Furthermore, the calculation results of current building codes are introduced into the learning of input-output functional mapping as potential inputs to improve prediction accuracy and to suggest improvements to these building codes. The results show that introducing European building codes significantly improves the prediction accuracy to a level that is significantly above that achievable using the initial parameters alone. In addition, the results highlight that improvements of particular building codes are relevant to different parameter combinations. Moreover, suggestions for future modifications of European building codes were brought out.