ASCE Press- Scouring at Bridge Piers Using Artificial Intelligence Models

Advanced AI techniques—including Gene-Expression Programming, Evolutionary Polynomial Regression, Model Tree, and Multivariate Adaptive Regression Spline—are benchmarked for predicting local scour depth at bridge piers. The work highlights the value of physical consistency and interpretability over traditional black-box methods. Scouring at Bridge Piers Using Artificial Intelligence Models: Implementation and Prediction evaluates the effectiveness of various Artificial Intelligence (AI) models—such as Gene-Expression Programming (GEP), Evolutionary Polynomial Regression (EPR), Model Tree (MT), and Multivariate Adaptive Regression Spline (MARS)—in predicting local scour depth at bridge piers, emphasizing their physical consistency and interpretability compared to traditional methods. The author highlights the limitations of black-box AI models and aims to improve the empirical understanding of scouring data through advanced statistical analysis. Topics include · Methodologies to estimate scouring at bridge piers, · Effective parameters, · AI models and their setting parameters, and · Practical examples of AI models in scour depth prediction. Mohammad Najafzadeh has compiled this book as an innovative resource for bridge designers and professionals in field investigations and consultant engineering, such as hydraulic and transportation engineers, as well as professors and graduate students who seek to explore new potential for scouring at bridge piers using empirical equations and AI models to generate precise estimations of scour depth.