Chapter 1 Damage Assessment of Steel Structures Using Multi-Autoregressive Model Chin-Hsiung Loh and Chun-Kai Chan Abstract For application in operational modal analysis considering simultaneously the temporal and spatial response data of multi-channel measurements, the multivariate-autoregressive (MV-AR) model was used. The parameters of MV-AR model are estimated by using the least squares method via the implementation of the QR factorization as an essential numerical tool and are used to extract the structural damage sensitive features. These parameters are used to develop the Vectors of autoregressive model and Mahalanobis distance, and then to identify the damage features and damage locations. Verification of the proposed method using a series of white noise response data of a steel structure is demonstrated. This method is thus very effective for damage detection in case of ambient vibrations dealing with output-only modal analysis. In addition, comparisons and discussions on the proposed method with other methods, such as stochastic subspace identification and wavelet-based energy index, are also presented. Keywords Damage detection • Multi-autoregressive model • Damage sensitive feature • SSI-COV • Wavelet-based energy index 1.1 Introduction Structural system identification and damage detection have received more and more attention in the field of civil engineering. Through monitoring data on structures a quantity of information can be obtained. A well known classification for damage identification methods can be defined four levels: (Level-1) Determination or detection that damage is present in the structure, (Level-2) Determination of the geometric location of the damage, (Level-3) Quantification of the severity of the damage, (Level-4) Prediction of the remaining service life of the structure. Generally, detection is performed by pattern recognition methods or Novelty detection [16], and the key issue for inverse methods is the damage location identification. Once the damage is located, it may be parameterized with a limited set of parameters and quantification. One of the efficiently and accurately monitoring techniques to all types of structural systems is the vibration-based damage detection. It is based on the principal that damage in a structure will alter the dynamic response of that structure and the selection of damage-sensitive features such as natural frequencies [12], displacement mode shapes [5], wavelet analysis of dynamic signals [7] will be concerned. In the viewpoint of global monitoring, system identification techniques extract natural frequencies, damping ratios, and mode shapes of a structure using acceleration data [3, 8]. But these dynamic features are not sensitive to damage. Fundamentally, feature extraction can also be based on fitting some model, either physics-based [Moaveni et al., 2008] or data-driven [13], to the measured system response data. Time-series analysis based on the use of autoregressive (AR) models have been extensively used in the SHM process as a feature extraction technique and also applied to damage detection [10, 11]. The algorithm is based on the premise that structural damage will change the vibration response of the structure. In this study, damage identification on the seismic response of two steel structures is examined. By using the response data of the two structures from white noise excitation between a series of earthquake excitations back to back from the shaking table tests, the multivariate signal processing techniques are used to extract the dynamic features directly from response measurements. The state of damage severity as well as the damage location through the developed novel damage detection algorithms is proposed. C.-H. Loh ( ) • C.-K. Chan Department of Civil Engineering, National Taiwan University, Taipei 10617, Taiwan e-mail: lohc0220@ccms.ntu.edu.tw; r98521205@ntu.edu.tw © The Society for Experimental Mechanics, Inc. 2016 S. Pakzad, C. Juan (eds.), Dynamics of Civil Structures, Volume 2, Conference Proceedings of the Society for Experimental Mechanics Series, DOI 10.1007/978-3-319-29751-4_1 1
RkJQdWJsaXNoZXIy MTMzNzEzMQ==