A novel load-dependent sensor placement method for model updating based on time-dependent reliability optimization considering multi-source uncertainties

An effective sensor network with an appropriate sensor configuration is the first step of model updating to obtain the actual structural response. However, sensor placements based on inherent structural characteristics (such as mode shapes) alone or their optimizations only with deterministic data are unlikely to provide very good results. Therefore, using a non-probabilistic theory to characterize the uncertainty in the uncertainty propagation process for model updating, this study proposes a time-dependent, reliability-based method for the optimal load-dependent sensor placement considering multi-source uncertainties. Due to the limitations of the uncertain parameters obtained using probabilistic or statistical methods, the uncertainties tackled in this study that includes those from structural properties and measurement processes are regarded as interval variables. Using the first-passage theory in the overall time history, different crossing situations of the reduced time history responses (that is, the modal coordinates) with respect to the full ones are constructed. The difference between the modal coordinates of the reduced and the full models is defined as the objective of the optimization, which indicates the matching level. Based on the time-dependent reliability-based index and the errors of deterministic modal coordinates between the reduced and full models, the multi-objective optimization is solved using NSGA-II. A detailed flowchart of the proposed method is given, and its effectiveness is verified by two simulated engineering examples for model updating.

Automated summary

This study proposes a time-dependent, reliability-based method for optimal load-dependent sensor placement considering multi-source uncertainties. Uncertainties from structural properties and measurement processes are regarded as interval variables in the study, and an optimization approach using NSGA-II is used to match the modal coordinates of reduced and full models.