Research on Seismic Performance of Bridge Piers Based on Machine Learning

Abstract

The shear capacity of bridge pier components serves as a core control indicator that determines their seismic performance and ensures the seismic safety of bridge engineering. In the existing database, short pier specimens with a shear span ratio of less than 2.0 account for 61.3% of the samples, and specimens with a high axial compression ratio greater than 0.4 account for 68.5% of the samples. Relevant experimental data are lacking regarding irregular-shaped bridge piers of long-span bridges in high-intensity areas and shear failure in plastic hinge regions under cyclic loading from strong earthquakes. The prediction accuracy of traditional formulas is generally lower than that of machine learning models.