DNN-Based Estimation of the Maximum Lateral Flange Moments of Horizontally Curved I-Girder Bridges

Horizontally curved I-girder bridges are known to be complex. Bending and torsion forces imposed on the owing their shapes with initial curvatures. This is a combination of pure warping forces. The horizontally significantly affected by behavior, which decreases bending rigidity its member. To investigate behavior finite element analysis (FEA) must performed. In this study, an FEA was performed torsional bridge, structural response database obtained. Based database, least absolute shrinkage selection operator employed select features affecting behavior. Subsequently, deep neural network models were trained selected for input layer maximum lateral flange moment data output layer. Several constructed compared according number hidden layers neurons, model highest performance proposed. Finally, it confirmed that estimated moments computed proposed showed good correlation results.