FAME–SSA : a Fast Adaptive Multivariate Decomposition Method for unsupervised damage detection in built infrastructures

Tools

Lakhadive, Mehulkumar and Sharma, Anshu and Bhowmik, Basuraj (2026) FAME–SSA : a Fast Adaptive Multivariate Decomposition Method for unsupervised damage detection in built infrastructures. Journal of Sound and Vibration, 629. 119676. ISSN 0022-460X (https://doi.org/10.1016/j.jsv.2026.119676)

[thumbnail of Mehulkumar-etal-JSV-2026-a-Fast-Adaptive-Multivariate-Decomposition-Method-for-unsupervised-damage-detection]
Preview
 Text. Filename: Mehulkumar-etal-JSV-2026-a-Fast-Adaptive-Multivariate-Decomposition-Method-for-unsupervised-damage-detection.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (15MB)| Preview

Abstract

Structural Health Monitoring (SHM) techniques are continuously challenged by the complexities and noise inherent in large-scale multidimensional data. Empirical Mode Decomposition (EMD) is effective for non-stationary signals but struggles with multichannel data. Multivariate EMD (MEMD) addresses this but still suffers from noise sensitivity, mode mixing, and incomplete frequency extraction. Fast and Adaptive Multivariate EMD (FA-MVEMD) improves on MEMD by using intelligent strategies to enhance accuracy and performance. This study introduces a novel methodology for modal identification and damage detection: Fast and Adaptive Multivariate Empirical Singular Spectrum Analysis (FAME-SSA). By combining the adaptive decomposition capability of FA-MVEMD with the trend extraction and noise separation strengths of Singular Spectrum Analysis (SSA), the proposed approach improves the accuracy and robustness of feature extraction from structural responses. A key innovation of FAME-SSA is the application of Hotelling’s T² and Squared Prediction Error (SPE) statistics for damage detection. The method is validated through extensive numerical simulations, experimental data from a wind turbine structure, and real-world SHM data from the Lysefjord Bridge. The results demonstrate that FAME-SSA outperforms conventional methods, making it a promising tool for real-time SHM in complex and noisy environments under challenging conditions.

ORCID iDs

Lakhadive, Mehulkumar, Sharma, Anshu ORCID logoORCID: https://orcid.org/0009-0004-4945-9821 and Bhowmik, Basuraj ORCID logoORCID: https://orcid.org/0000-0001-7782-513X;
CORE (COnnecting REpositories)