Abstract
The structural integrity of buildings and infrastructures can be affected by either environmental conditions or unforeseen external actions. In order to efficiently detect damage, intended as an irreversible degradation of the structural stiffness, many identification algorithms have been proposed in the literature. Nevertheless, a crucial aspect to accurately estimate and locate such damage pertains to the configuration of the deployed structural health monitoring (SHM) system. In addressing this goal, a framework is here proposed for the optimal design of sensor networks, in terms of number, type and spatial deployment of the sensors. The rationale of the method is to simultaneously maximize the information associated with the measurements, and minimize the total cost of the experimental setup; the overarching goal thus lies in the maximization of the information per unit cost, for the efficient allocation of resources. The value of the SHM system is quantified through the Shannon information gain between the a-priori knowledge of the mechanical properties and the values estimated, on the basis of measurements. The types of sensors contained into the overall SHM mix largely affects the estimation accuracy since, as a rule of thumb, the higher the sensor cost, the higher the signal-to-noise ratio and, therefore, the better the attainable estimation. In order to tackle the aforementioned multi objective optimization problem and to derive the associated Pareto front, a-posteriori solution methods relying on evolutionary algorithms are adopted. The proposed method is applied to a shear-type structure, namely the Pirelli tower in Milan, and the relevant multi-criteria optimization solutions are presented.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have