Structural Time-Varying Parameter Analysis Based on Adaptive System Identification Method Using Strong Motion Records

To diagnose and predict the seismic damage of buildings under the excitation of strong earthquake, the structural time-varying parameters are identified based on the strong motion records by using on-line system identification method. As one example, the natural frequency and damping ratio of a 5-story reinforcement concrete structure are obtained from its strong motion records. The results show that the time-varying characteristics of structure can be predicted accurately by using adaptive forgetting through multiple models, and the structural parameters of buildings before and after damage can be obtained from the identified time-varying parameters. It is also found that the natural frequency decreases gradually until it reaches the lowest point value, and then increases until the end of the earthquake, but it can not be recovered to the initial value, however, the damping ratio changes irregularly during strong earthquake. The performance of structures could be recovered to a certain extent after the earthquake compared with the most dangerous state when the response reaches the maximum value. The results can be referenced by the damage detection and diagnosis.