Tracking modal parameters of building structures from experimental studies and earthquake response measurements

Since tracking modal parameters and estimating the structural current state from building seismic response measurement can provide useful information for building safety assessment, online or recursive identification techniques need to be developed and implemented on building seismic response monitoring. In this article, two recursive subspace identification algorithms are developed. One is the recursive subspace identification with BonaFide LQ renewing algorithm (recursive subspace identification BonaFide) incorporated with moving window technique to identify modal parameters at each time instant during the earthquake excitation. Besides, recursive subspace identification based on matrix inversion lemma algorithm (recursive subspace identification Inversion) is also used and keeps the same data window with recursive subspace identification BonaFide only for the initial identification and then appends the following data points to identify modal parameters for the whole excitation time history. A forgetting factor is introduced to emphasize the latest state of system in this method. Two different sets of building seismic response data are used to verify the applicability of the proposed methods. One is the three-story steel structure with abrupt change of story stiffness shaking table test in laboratory, the other is a series of seismic response of a four-story-reinforced concrete elementary school building, including several large seismic events. It is concluded that both methods can detect the time when the floor stiffness was changed but recursive subspace identification Inversion with forgetting factor can provide more accurate estimation of the change of stiffness.