Automated operational modal analysis for civil engineering structures with compressed measurements

Introducing compressive sensing (CS) to modal analysis enables modal identification approaches to process compressed measurements, which enhances the efficiency in both data transmission and analysis for long-term monitoring of civil engineering structures. Nevertheless, the existing methods normally require users' judgment in mode selection, and thus they cannot realize automated modal analysis. To bridge this gap, this study proposes an automated CS-based operational modal analysis (OMA) method, namely Automated Sparse Decomposition (ASD). A mode selection strategy is proposed to gradually eliminate noise and false modes and thus realize automated modal identification based on compressed measurements. A search strategy with both extensive and intensive search stages is proposed to realize higher frequency search resolution with low computational costs. Additionally, an automated mode shape identification process is designed to enhance accuracy. To demonstrate the effectiveness of the proposed method, experimental data on an offshore wind turbine model and real-world monitoring data from an in-service cable-stayed bridge are analyzed. The results show that the identification accuracy of the proposed ASD is comparable to state-of-art CS-based methods. Due to its superior performance in both efficiency and accuracy, the proposed method has the potential to become a promising online modal analysis tool.