Extended Bouc-Wen model identification using shaking table test data of ageing RC bridge piers

In scientific research or engineering design, analysis with a large amount of repeating (e.g. incremental dynamic analysis, fragility analysis) is needed. In this process, low-order models require less computational cost compared with sophisticated finite element models. This paper explores the feasibility of identifying an extended Bouc-Wen model with input ground motion and structural responses of displacement and acceleration recorded in the shaking table tests. The parameters to be estimated are selected based on their sensitivity. A boundary of the unknown parameters is designated based on their physical meaning or empirical values. Genetic algorithm is applied to search for the optimal fitting of the extended Bouc-Wen model. The simulated results are compared with the experimental results in terms of hysteresis and time-varying stiffness/frequency during the test.