comparison of fragility curves based on soft computing and analytical methods on horizontally curved bridges

Today, a wide range of tools and software are available for nonlinear dynamic analysis as a comprehensive method for estimation of seismic demand in highway bridges. generally, nonlinear structural analysis methods based on the structural mechanic and mathematical methods are using in the estimation of the structural vulnerabilities. the soft computing(sc) components include multiple main branches and many techniques inspired by the activity of the human brain that has proven useful in solving complex problems. among sc methods, the artificial neural network (ann) is widely is employed in engineering optimization problems. in the recent decades, a significant number of horizontally-curved steelbridges have been built in urban areas. the comparison of these bridges with straight bridges indicates fewer studies on the seismic vulnerability of these bridges. figure 1 represents the structure of neural network using structural characteristics, earthquake ground motion records and engineering demand parameters.