Coupling physics-based models with wireless sensor networks for structural health monitoring

In wireless structural health monitoring (SHM) systems, the over-the-air transmission of large data sets – apart from being unreliable – may drastically reduce the power autonomy of wireless sensor nodes, thus raising the need for continuous power supply, which effectively cancels the “wireless” nature of the sensor nodes. To address the aforementioned shortcom- ings, this paper reports on advantageously utilizing the embedded computing capabilities of wireless sensor nodes for extracting information on the structural properties on board, i.e. with- out resorting to centralized data acquisition. Drawing from emerging paradigms associated with the digitalization of physical processes, e.g. using digital twins, the embedded physics-based modeling concept presented herein couples physics-based models of structures with wireless sensor networks. In particular, physics-based models, which are frequently adopted in SHM for mapping the outcome of (typically model-agnostic) data analysis to structural behavior phenom- ena perceivable by structural engineers, can be embedded into wireless sensor nodes, facilitat- ing advanced local autonomous data analysis on board the sensor nodes. The embedded phys- ics-based modeling concept is validated via simulations, showcasing the capability of the em- bedded physics-based models to yield information on the structural properties. The results of the proposed concept are expected to align SHM practices with digitalization paradigms that form the backbone of Industry 4.0.

Options

Coupling physics-based models with wireless sensor networks for structural health monitoring