Vibration-based Structural Health Monitoring for beams under temperature variation using integrated acceleration, velocity and displacement features

• João Carvalho UFJF
• Jessé Mazzetti UFJF
• Tales Boratto UFJF
• Heder Bernardino UFJF
• Alexandre Cury UFJF
• Leonardo Goliatt UFJF

DOI: https://doi.org/10.5753/eniac.2026.31319

Resumo

Damage classification is important in structural health monitoring (SHM), yet achieving high reliability remains challenging due to complex, multi-domain vibration signals. This work focuses on feature engineering, combining derived features from the acceleration, velocity, and displacement domains, obtained by integrating accelerometer data from two datasets. Thus, a framework is proposed to classify levels of damage in beams subjected to different temperatures, through a comparative analysis involving four supervised algorithms: Decision Tree, Random Forest, XGBoost, and Logistic Regression. Finally this approach was effective, since XGBoost achieved over 95% performance on key evaluation metrics with feature combinations from the three physical domains.

Referências

Ahmetoglu, H. and Das, R. (2022). A comprehensive review on detection of cyberattacks: Data sets, methods, challenges, and future research directions. Internet of Things, 20:100615.

Ali, J., Khan, R., Ahmad, N., and Maqsood, I. (2012). Random forests and decision trees. International Journal of Computer Science Issues(IJCSI), 9.

Alves, V. and Cury, A. (2023). An automated vibration-based structural damage localization strategy using filter-type feature selection. Mechanical Systems and Signal Processing, 190:110145.

Boratto, T., Lima Fonseca, D., Soares Bernardino, H., Borges, A., Cury, A., and Goliatt, L. (2025). On the use of spectral data from smartphone accelerometer signals and constituent material for the identification of damaged walls. Rev. Inform. Teór. Apl., 32(2):52–63.

Brandt, A. and Brincker, R. (2014). Integrating time signals in frequency domain – comparison with time domain integration. Measurement, 58:511–519.

Cramer, J. S. (2002). The origins of logistic regression. Tinbergen Institute Working Paper 2002-119/4, Tinbergen Institute.

Gilbert, H. B., Celik, O., and O’Malley, M. K. (2010). Long-term double integration of acceleration for position sensing and frequency domain system identification. In Proceedings of the IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), pages 453–458, Montréal, Canada.

Huang, D., Du, J., Zhang, L., Zhao, D., Deng, L., and Chen, Y. (2014). A wavelet-based algorithm for numerical integration on vibration acceleration measurement data. Journal of Vibroengineering, 16(7):3469–3479.

Ibrahem Ahmed Osman, A., Najah Ahmed, A., Chow, M. F., Feng Huang, Y., and El-Shafie, A. (2021). Extreme gradient boosting (xgboost) model to predict the ground-water levels in selangor malaysia. Ain Shams Engineering Journal, 12(2):1545–1556.

Ma, Z., Choi, J., and Sohn, H. (2023). Structural displacement sensing techniques for civil infrastructure: A review. Journal of Infrastructure Intelligence and Resilience, 2(3):100041.

Morales, F. A. O., Cury, A. A., and Peixoto, R. A. F. (2018). Analysis of thermal and damage effects over structural modal parameters.

Qiu, Y., Ahmed, B., Abueidda, D. W., El-Sekelly, W., de Soto, B. G., Abdoun, T., Ji, H., Qiu, J., and Mobasher, M. E. (2026). Damage identification for bridges using machine learning: Development and application to kw51 bridge. Digital Engineering, 10:100103.

Sekiya, H., Kimura, K., and Miki, C. (2016). Technique for determining bridge displacement response using mems accelerometers. Sensors, 16(2):257.

Taylor, J. (2003). The Vibration Analysis Handbook: A Practical Guide for Solving Rotating Machinery Problems. VCI.

Yang, Y., Zhao, Y., and Kang, D. (2016). Integration on acceleration signals by adjusting with envelopes. Journal of Measurements in Engineering, 4(2):117–121.

Zheng, W., Dan, D., Cheng, W., and Xia, Y. (2019). Real-time dynamic displacement monitoring with double integration of acceleration based on recursive least squares method. Measurement, 141:460–471.