A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Artificial neural network and machine learning models for predicting the lateral cyclic response of post-tensioned base rocking steel bridge piers
Post-tensioned rocking steel bridge piers are round tubes with welded circular base plates. A post-tensioned (PT) tendon or gravitation loads are utilized to compress the structure's foundation. In this research, an artificial neural network (ANN) model and six machine learning (ML) techniques were considered to determine which would be most effective at predicting the lateral cyclic response of PT base rocking steel bridge piers. These ML techniques included linear regression, support vector regression, decision trees, K-nearest neighbors, random forests, and extreme gradient boosting. Factors such as tendon cross-sectional area, initial post-tensioning ratio, dead-load ratio, base plate thickness, and base plate extension were taken into account. Column diameter, column diameter-to-thickness ratio, column height-to-diameter ratio, and column height were also taken into account. The study takes into account the residual drift of the columns, the shortening of the columns, the ratio of the degraded stiffness to the starting stiffness, the maximum lateral strength to the uplift force, and the lateral strength reduction ratio as response factors. The proposed strategy was tested using a number of statistical measures, such as R-squared (R2), root mean square error (RMSE), and mean absolute error (MAE). When compared to the other models under consideration, the random forest based model is recommended due to its superior prediction performance, as measured by a greater coefficient of determination and a lower error estimate.
Artificial neural network and machine learning models for predicting the lateral cyclic response of post-tensioned base rocking steel bridge piers
Post-tensioned rocking steel bridge piers are round tubes with welded circular base plates. A post-tensioned (PT) tendon or gravitation loads are utilized to compress the structure's foundation. In this research, an artificial neural network (ANN) model and six machine learning (ML) techniques were considered to determine which would be most effective at predicting the lateral cyclic response of PT base rocking steel bridge piers. These ML techniques included linear regression, support vector regression, decision trees, K-nearest neighbors, random forests, and extreme gradient boosting. Factors such as tendon cross-sectional area, initial post-tensioning ratio, dead-load ratio, base plate thickness, and base plate extension were taken into account. Column diameter, column diameter-to-thickness ratio, column height-to-diameter ratio, and column height were also taken into account. The study takes into account the residual drift of the columns, the shortening of the columns, the ratio of the degraded stiffness to the starting stiffness, the maximum lateral strength to the uplift force, and the lateral strength reduction ratio as response factors. The proposed strategy was tested using a number of statistical measures, such as R-squared (R2), root mean square error (RMSE), and mean absolute error (MAE). When compared to the other models under consideration, the random forest based model is recommended due to its superior prediction performance, as measured by a greater coefficient of determination and a lower error estimate.
Artificial neural network and machine learning models for predicting the lateral cyclic response of post-tensioned base rocking steel bridge piers
Asian J Civ Eng
Nabizadeh, Elham (author) / Parghi, Anant (author)
Asian Journal of Civil Engineering ; 25 ; 511-523
2024-01-01
13 pages
Article (Journal)
Electronic Resource
English
Characterization of the Lateral Response of Base Rocking Steel Bridge Piers
| Springer Verlag | 2022
Biaxial testing of unbonded post‐tensioned rocking bridge piers with external replacable dissipaters
| Online Contents | 2011