Stochastic modal analysis for a real offshore platform
Date of Original Version
The stochastic subspace identification (SSI) method, which is based on a first-order state-space stochastic model, has been utilized for system identifications in various engineering applications. This paper is to study the difference of the projection-driven (SSI) and covariance-driven (SRA) methods using measured data from a real offshore platform. In theory, the SSI and SRA methods are similar in many ways. Firstly, they are modeled by the same state-space equations, and utilized exactly the same noisy response data. Secondly, they both estimate the modal parameters through a system realization algorithm that involves the truncated singular value decomposition of a kernel matrix. In practical applications, the sole difference of the two methods is related to the computation of either projection or covariance from the same noisy raw data, noting that both the projection and covariance operations are aimed to cancel out the (uncorrelated) noise. For the operational modal analysis of a real offshore structure, the projection-driven method was recommended over the covariance-driven method when these methods were employed together with a stability diagram to identify the "true" system poles, because multiple spurious poles could be stabilized in its stabilization diagram at the frequency locations that are near to the true poles associated with strongly excited modes while implementing the covariance-driven SRA method.
Publication Title, e.g., Journal
6th International Operational Modal Analysis Conference, IOMAC 2015
Liu, Fushun, Huajun Li, and Sau-Lon J. Hu. "Stochastic modal analysis for a real offshore platform." 6th International Operational Modal Analysis Conference, IOMAC 2015 (2015). https://digitalcommons.uri.edu/oce_facpubs/256