Structural damage localization using wavelet-based silhouette statistics

This paper introduces a new methodology for classifying and localizing structural damage in a truss structure. The application of wavelet analysis along with signal classification techniques in engineering problems allows us to discover novel characteristics that can be used for the diagnosis and classification of structural defects. This study exploits the data discriminating capability of silhouette statistics, which is eventually combined with the wavelet-based vertical energy threshold technique for the purpose of extracting damage-sensitive features and clustering signals of the same class. This threshold technique allows us to first obtain a suitable subset of the extracted or modified features of our data, i.e. good predictor sets should contain features that are strongly correlated to the characteristics of the data without considering the classification method used, although each of these features should be as uncorrelated with each other as possible. The silhouette statistics have been used to assess the quality of clustering by measuring how well an object is assigned to its corresponding cluster. We use this concept for the discriminant power function used in this paper. The simulation results of damage detection in a truss structure show that the approach proposed in this study can be successfully applied for locating both open- and breathing-type damage even in the presence of a considerable amount of process and measurement noise. Finally, a typical data mining tool such as classification and regression tree (CART) quantitatively evaluates the performance of the damage localization results in terms of the misclassification error.