A novel and near-automatic mode tracking algorithm for civil infrastructure

Abstract

Robustly and accurately tracking the evolution of modal parameters of large civil infrastructure is key for learning about their operational behaviour and performing modal-based structural health monitoring. This work proposes a new, two-step, mode tracking (MT) algorithm capable of automatically identifying which structural modes are to be tracked, tracking these modes even if they are subject to significant changes, and doing so independent of the (automatic) operational modal analysis (A)OMA method used for the modal identification. The MT algorithm is thoroughly tested using numerical benchmarking data, showing that the proposed algorithm can successfully and repeatedly identify and track structural modes, even if the modal detections are irregular, that the modal properties undergo cyclic variations, and that they are subject to abrupt changes. The MT algorithm is applied to two complex real-world cases, the Hardanger suspension bridge and the Bergsøysund floating pontoon bridge. It can identify and track all the expected modes in these two challenging cases.