A Comparative Study of AI Algorithms for Long -Term SHM in Concrete High -Rise Buildings

The long -term structural health monitoring (SHM) of concrete high -rise buildings is critical for ensuring safety, functionality, and resilience against environmental and operational stresses. With the increasing complexity of sensor data and the demand for real -time decision -making, artificial intelligence (AI) has emerged as a transformative tool in SHM systems. This study presents a comprehensive comparative analysis of several AI algorithms —namely Support Vector Machines (SVM), Random Forests (RF), Convolutional Neural Networks (CNN), and Long Short-Term Memory networks (LSTM) —applied to long -term monitoring of high -rise concrete structures. A benchmark dataset derived from real -world SHM installations, including vibration signals, strain responses, and environmental parameters, is used to train and evaluate these models. The performance of each algorithm is assessed based on accuracy, robustness to noise, computational efficiency, and adaptability to temporal changes. The results reveal that while deep learning models such as LSTM and CNN exhibit superior accuracy in capturing complex structural patterns over time, traditional machine learning methods like RF offer faster inference and lower computational overhead. The findings provide actionable insights for SHM practitioners seeking to deploy scalable, data -driven monitoring frameworks for the long -term maintenance of urban infrastructure.

Keywords:

Structural health monitoring (SHM) , Concrete high -rise buildings , Artificial intelligence (AI) , Machine learning , Deep learning , Support vector machine (SVM) , Random Forest (RF) , Convolutional neural network (CNN) , Long short -term memory (LSTM) , Long -term monitoring