Modal Parameter Determination of Stay Cable with an Improved Algorithm Based on Stochastic Subspace Identification

Abstract : The modal parameter determination of stay cable is generally a critical issue for either designing or monitoring cable-stayed bridges. Stochastic subspace identification (SSI) has been proved in recent years to be an excellent tool for obtaining all the modal parameters together in the cases where only the output signals are available. Nevertheless, its feasibility and accuracy in practical applications remain to be extensively verified. Especially for the situations where the frequency content of excitation is narrowly banded and thus far away from the white-noise assumption underneath the theoretical derivation of SSI, the problem of superfluously indentifying numerous deceptive modes usually occurs. Attempting to crack this difficulty, an improved algorithm based on SSI is developed in the current work for performing the modal parameter determination of stay cables. This study adopts the covariance type of SSI, usually with better stability and computation efficiency, to establish an effective methodology for extensively identifying the modal parameters of stay cables. Several details of choosing the computational parameters in performing SSI are first discussed, followed by proposing an alternative stabilization diagram such that most modal parameters of cable can stand out and then imposing appropriate sifting criteria to extract the most reliable modal parameters.