Automated and Model-Free Bridge Damage Indicators with Simultaneous Multiparameter Modal Anomaly Detection

Area of Science:

• Structural Engineering
• Vibration Analysis
• Structural Health Monitoring (SHM)

Background:

• Structural damage identification is crucial for infrastructure safety.
• Vibration-based methods offer non-destructive assessment capabilities.
• Automated analysis reduces human intervention and enhances efficiency.

Purpose of the Study:

• To develop a simultaneous modal parameter anomaly detection paradigm for structural damage identification.
• To automate modal parameter extraction and anomaly recognition using data-driven techniques.
• To evaluate the performance of the proposed method on real-world bridge structures.

Main Methods:

• System Realization Using Information Matrix (SRIM) for state matrix identification.
• Clustering of stable modal poles from stability diagrams.
• Gaussian distribution-based anomaly detection on frequency and damping ratios.
• Boolean operators for merging univariate anomalies.

Main Results:

• Successful automated modal analysis and anomaly recognition with minimal human supervision.
• Demonstrated realistic performance on two distinct bridge structures (reinforced concrete and steel arch).
• Anomaly detection performance is sensitive to threshold selection and information retrieval metrics.

Conclusions:

• The proposed methodology is effective for data-driven, scalable, and fully autonomous SHM.
• The automated approach significantly reduces the need for human intervention in damage identification.
• The method provides a robust framework for real-time structural health assessment.