A Mobile Sensing Framework for Bridge Modal Identification through an Inverse Problem Solution Procedure and...

Area of Science:

• Structural Health Monitoring
• Smart Infrastructure
• Transportation Engineering

Background:

• Smart infrastructure development necessitates cost-effective bridge monitoring systems.
• Traditional fixed sensor systems are expensive and may not cover all critical areas.
• Vehicle-mounted sensors offer a potential solution for reduced monitoring costs.

Purpose of the Study:

• To present an innovative framework for determining bridge response and identifying modal characteristics.
• To utilize only accelerometer data from vehicles passing over the bridge.
• To reduce the cost of bridge monitoring systems.

Main Methods:

• Determining virtual fixed node responses using vehicle axle acceleration as input.
• Employing inverse problem solutions with linear and cubic spline functions for initial response estimation.
• Utilizing a moving-window autoregressive model with exogenous inputs (ARX) for signal prediction in error regions.
• Integrating Singular Value Decomposition (SVD) and Frequency Domain Decomposition (FDD) for modal identification.

Main Results:

• The proposed framework accurately determines bridge acceleration and displacement responses.
• The ARX model effectively predicts responses in regions with large errors.
• Integration of SVD and FDD successfully identifies bridge mode shapes and natural frequencies.
• The method demonstrates high accuracy in identifying the three main bridge modes under various conditions.

Conclusions:

• The developed framework offers a cost-effective and accurate method for bridge structural health monitoring.
• Utilizing vehicle-mounted accelerometers is a viable alternative to traditional fixed sensor systems.
• The proposed approach shows robustness against ambient noise and variations in vehicle parameters.