Bridge modal identification using acceleration measurements within moving vehicles

• Published in: Mechanical systems and signal processing 2020-07, Vol.141, p.106733, Article 106733
• Main Authors: Sadeghi Eshkevari, Soheil, Matarazzo, Thomas J., Pakzad, Shamim N.
• Format: Article
• Language: English
• Subjects: Algorithms; Blind source separation; Damping ratio; Data collection; Dynamic response; Empirical analysis; Frequency response function; Identification; Identification methods; Methods; Mobile sensing; Modal identification; Numerical methods; Output-only; Resonant frequencies; Sensors; Signal processing; Structural health monitoring; Surface roughness; Suspension systems; System identification; Traffic volume; Transfer functions; Vehicles
• ISSN: 0888-3270; 1096-1216
• DOI: 10.1016/j.ymssp.2020.106733

•This study is the first to propose end-to-end pipelines for comprehensive modal identification of a bridge using moving sensors within vehicles.•Both proposed methods are successful to remove vehicle suspension effects and roughness-induced vibrations from the collected data within the vehicle cabin in order to extract bridge dynamic response.•The study indicates that the FRF-based method is able to identify the bridge with high accuracy as well as an acceptable estimation of the roughness profile.•The EEMD-based method does not need any a priori information of the vehicle, leading a practically more flexible platform.•In a numerical 2D case study, first four modes are fully identified using the proposed pipelines. MAC values are all above 0.94. Natural frequencies and damping ratios are also estimated accurately for the majority of cases. Vehicles commuting over bridge structures respond dynamically to the bridge’s vibrations. An acceleration signal collected within a moving vehicle contains a trace of the bridge’s structural response, but also includes other sources such as the vehicle suspension system and surface roughness-induced vibrations. This paper introduces two general methods for the bridge system identification using data exclusively collected by a network of moving vehicles. The contributions of the vehicle suspension system are removed by deconvolving the vehicle response in frequency domain. The first approach utilizes the vehicle transfer function, and the second uses ensemble empirical modal decomposition (EEMD). Next, roughness-induced vibrations are extracted through a novel application of second-order blind identification (SOBI) method. After these two processes, the resulting signal is equivalent to the readings of mobile sensors that scan the bridge’s dynamic response. Structural modal identification using mobile sensor data has been recently made possible with the extended structural modal identification using expectation maximization (STRIDEX) algorithm. The processed mobile sensor data is analyzed using STRIDEX to identify the modal properties of the bridge. The performance of the methods are validated on numerical case studies of a long single-span bridge with a network of moving vehicles collecting data while in motion. The analyses consider three road surface roughness patterns. Results show that for long-span bridges with medium- to high-ongoing traffic volume, the proposed algorithms are successful in extracting pure bridge