28 February 2024 in Structural Health Monitoring
Over the past decades, Vibration-Based Methods (VBMs) have consistently exhibited exceptional effectiveness in the field of Structural Health Monitoring when it comes to assessing structural damage in both civil and mechanical structures. Recently, the progress made in data-driven strategies for localizing structural damage through the VBMs has resulted in substantial benefits. These advanced strategies not only enable an efficient decision-making process but also provide precise identification of repair locations for monitored structures. Importantly, they achieve this without the inherent complexity and computational burden typically associated with traditional model-based methods. In this study, an innovative data-driven method for localizing and quantifying structural damage is proposed. The method is developed on the basis of the principles of Linear Discriminant Analysis (LDA) and a newly devised modeling strategy that utilizes the power cepstral coefficients extracted from the structural acceleration response. The developed LDA model is able to highlight the local structural characteristics (i.e., mode shape-related information) embedded in the cepstral coefficients in the LDA latent space. Based on the highlighted local characteristics of the cepstral coefficients, a statistical pattern recognition strategy is proposed to effectively conduct the quantification and localization of structural damage. The proposed damage localization method is designed to function in a completely unsupervised-learning manner, which eliminates the requirement for the model to have access to any prior knowledge or information regarding structural damage during the training phase. The proposed method has been validated by both numerical simulations and experimental data.
24 February 2025 in Journal of Earthquake and Tsunami
1 December 2025 in Advances in Structural Engineering
The accuracy and practicality of structural parameter identification remain critical challenges in Structural Health Monitoring (SHM), as many existing modal updating techniques are limited by noise sensitivity, reliance on expert judgment, and complex system identification procedures. This study introduces a novel model updating method that redefines structural parameter identification process by utilizing Cepstral Coefficients (CCs) of structural responses, which provide a compact, robust, and noise-tolerant representation of dynamic characteristics. This study pioneers the use of response-cepstrum representations in structural model updating. Unlike conventional modal characteristics, the CCs are directly and efficiently extracted from time-domain responses through signal-processing techniques, avoiding the expertise-intensive and laborious modal identification process while significantly accelerating the analysis. Exploiting the unique characteristic representation of the CCs, a customized evolutionary optimization framework is developed based on an advanced variant of the Differential Evolution (DE) algorithm for parameter identification. This development is supported by a systematic investigation of multiple effective evolutionary optimization strategies to ensure robust and accurate matching of CCs between measured and simulated responses. The proposed method was validated through both numerical simulations and experimental tests, demonstrating strong performance in identification accuracy, computational efficiency, and robustness to measurement noise as well as parameter interdependence. This method offers an efficient and scalable alternative to modal updating and establishes a new direction for feature-driven, data-centric SHM systems.
5 March 2026 in Journal of Sound and Vibration