SEISMIC DAMAGE PREDICTION IN REINFORCED CONCRETE STRUCTURES USING MACHINE LEARNING-BASED MODELING

Abstract

Seismic damage assessment of reinforced concrete (RC) structures is a critical task in earthquake-prone regions to ensure structural safety and minimize economic losses. Conventional analytical and numerical methods often require extensive computational effort and detailed structural modeling. Recent advancements in machine learning (ML) provide data-driven alternatives capable of learning complex nonlinear relationships between seismic inputs and structural responses. This paper presents a machine learning–based framework for predicting seismic damage in RC structures using key structural and ground motion parameters. Supervised learning models are trained on simulated seismic response data to classify damage states effectively. The proposed approach improves prediction accuracy and reduces computational complexity compared to traditional methods. The results demonstrate that ML models can reliably estimate damage levels under varying seismic intensities. This study highlights the potential of intelligent systems for rapid post-earthquake damage assessment. The framework supports decision-making in structural design, retrofitting, and disaster mitigation planning.