Machine Learning Applications in Structural Engineering: Prediction Models for Moment-Rotation Characteristics in Boltless Steel Connections

Abstract

This article presents a study on the application of machine learning for predicting the moment-rotation behavior of boltless steel connections commonly used in pallet rack structures. Through extensive experimental data collection, Support Vector Machines (SVM) and Deep Learning (DL) models were developed to serve as predictive tools for these connections. The analysis demonstrates that these models enable engineers to accurately forecast structural characteristics, optimize boltless connection designs, and enhance the stability and functionality of pallet rack systems. The generalized model framework established here offers a robust foundation for future studies and design improvements, with SVM achieving the highest predictive accuracy among the models tested.