Real-time unified single- and multi-channel structural damage detection using recursive singular spectrum analysis

Abstract

A novel baseline free approach for continuous online damage detection of multi degree of freedom vibrating structures using Recursive Singular Spectral Analysis (RSSA) in conjunction with Time Varying Auto-Regressive Modeling (TVAR) is proposed in this paper. The acceleration data is used to obtain recursive proper orthogonal components online using rank-one perturbation method, followed by TVAR modeling of the first transformed response, to detect the change in the dynamic behavior of the vibrating system from its original state to contiguous linear /non-linear-states indicating damage. Most work to date deal with algorithms that require windowing of the gathered data that render them ineffective for online implementation. Algorithms focussed on mathematically consistent recursive techniques in a rigorous theoretical framework of structural damage detection is missing, that motivates the development of the present framework. The response from a single channel is provided as input to the algorithm in real time. The RSSA algorithm iterates the eigenvector and eigenvalue estimates for sample covariance matrices and new data point at each successive time instants. This eliminates the need for offline post processing and facilitates online damage detection especially when applied to streaming data without requiring any baseline data. Lower order TVAR models are applied on the transformed responses to improve detectability. Numerical simulations performed on a 5-dof nonlinear system and on an SDOF system modeled using a Duffing oscillator under white noise excitation data, with different levels of nonlinearity simulating the damage scenarios, demonstrate the robustness of the proposed algorithm. The method, further validated on results obtained from experiments performed on a cantilever beam subjected to earthquake excitation; a toy cart experiment model with springs attached to either side; demonstrate the efficacy of the proposed methodology as an appropriate candidate for real time, reference free structural health monitoring.

Scour Damage Detection and Structural Health Monitoring of a Laboratory-Scaled Bridge Using a Vibration Energy Harvesting Device

“…In most applications, it becomes necessary to remove the noise component prior to subsequent data analysis. Thus, the use of singular spectrum analysis (SSA) as a filter bank in noise reduction is well documented in the literature [33,34]. The method represents a potential alternative to the available filtering techniques based on eigen-decomposition on the Hankel covariance matrix obtained from a single channel of output data [34].…”

“…Thus, the use of singular spectrum analysis (SSA) as a filter bank in noise reduction is well documented in the literature [33,34]. The method represents a potential alternative to the available filtering techniques based on eigen-decomposition on the Hankel covariance matrix obtained from a single channel of output data [34]. The resulting time series can be reconstructed by using the principal components that correspond to the actual signal constituents, thereby leaving the random (or noise) component behind.…”

Hydropower Dam State and Its Foundation Soil Survey Using Industrial Seismic Oscillations

“…If, for some passive seismic methods that are based on the cross-correlation or convolution of ambient noise, a high level of microseisms is necessary [15][16][17]. For other methods that determine the natural frequency of a structure [3,10,18] or are based on spectral analysis of the natural ambient seismic noise [19][20][21], a high level of microseisms is a negative factor that impedes the data processing. Each of these methods has its own requirement for seismic noise level and frequency range, which is why coupling these methods in one experiment is difficult or sometimes impossible.…”

Structural Damage Diagnosis and Prediction Using Machine Learning and Deep Learning Models: Comprehensive Review of Advances<strong> </strong>

“…Due to the fact that artificial neural networks have various applications such as accurate prediction of complex material behavior, it could be applied for damage detection and structural integrities in corresponding multiple-variable problems. (Alves et al, 2015;Andrejiova, Grincova, & Marasova, 2019;Bhowmik, Krishnan, Hazra, & Pakrashi, 2019;Dia, Dieng, Gaillet, & Gning, 2019;Dorval, Meredieu, & Danjon, 2016;Egnew, Roueche, & Prevatt, 2018;Favillier et al, 2015;Kabir, Sadiq, & Tesfamariam, 2016;Kim, Hwang, & Jung, 2017;Noori Hoshyar, Samali, Liyanapathirana, & Taghavipour, 2019;Pérez-Ruiz et al, 2018;Wang et al, 2018; (Fakih, Mustapha, Tarraf, Ayoub, & Hamade, 2018;Froustey et al, 2017;J. Li & Zhang, 2016;Mayer & Mayer, 2019;Ovid'Ko, Sheinerman, Skiba, Krasnitiskiy, & Smirnov, 2015;Panettieri, Leclerc, Jumel, & Guitard, 2018;Y.…”