Organizers: Eloi Figueiredo and Ionut Moldovan, Lusófona University, Portugal
Scope of Session: Long-term structural health monitoring (SHM) has been mainly performed using two approaches: model- and data-based. A challenge in both approaches is to make the distinction between the variations of the structural response caused by damage and environmental or operational variability. Hybrid techniques for SHM that integrate model- and data-based approaches have emerged to complement the data measured by the monitoring system installed on the structure with data obtained from its numerical model, leading to both unsupervised and supervised machine learning strategies for damage identification. The hybrid approach to the SHM is still in its early development. Some of the challenges are related to the calibration of the numerical models such as producing reliable data. This Special Session intends to bring together publications involving the integration of numerical modeling, monitoring, and/or machine learning, in order to highlight the current capabilities and future trends. Besides theoretical contributions, the papers must address the practical engineering application of the techniques, using actual experimental or field monitoring data.
Keywords: machine learning, finite element, monitoring, SHM, detection, damage