Research Progress on Artificial Intelligence Methods in Structural Health Monitoring

Abstract

Structural health monitoring (SHM), as a core technology for ensuring the safe operation of infrastructure and engineering equipment, has become increasingly critical amidst the accelerating urbanization and increasing number of major projects. Traditional monitoring methods, limited by manual reliance and lack of real-time performance, are unable to adapt to the full lifecycle management needs of complex structures. Artificial intelligence, with its powerful data processing and pattern recognition capabilities, offers a breakthrough solution for SHM. This article systematically reviews the research progress of AI methods in SHM. It first explains the technical foundations of monitoring data acquisition and preprocessing, then analyzes the application mechanisms of classical machine learning and deep learning in damage identification. The success of these technologies is illustrated by case studies from various engineering fields, such as bridges and aerospace. Finally, the current technical bottlenecks are analyzed and future development directions are prospected. Research indicates that AI methods have made the leap from laboratory validation to engineering applications. Classical machine learning remains practical in small and medium-scale data scenarios, while deep learning demonstrates significant advantages in complex feature extraction. However, challenges such as data quality and algorithm generalization still require improvement. In the future, through the integration of multiple technologies and the deepening of engineering practice, AI will drive SHM towards precision, intelligence, and predictiveness, providing core support for the safety assurance of major projects.