Dynamic System Identification of Civil Structure using Genetic Algorithm

Abstract The aim of the contribution is creating of SDOF and simple MDOF models with included complex information of civil structures for dynamic analyses. The dynamic parameters of models are obtained as inverse problem solution. The inverse problem is defined as an optimization problem. The objective is to minimize the error index between measured response of the real structure and calculated response of the model. Unknown parameters are variables of optimization. The Genetic algorithm is used as an optimization procedure. Records of earthquake excitations and responding measured structural responses of multistory building are used as input and output of solved examples. The building structure in examples is modeled with help of linear and nonlinear SDOF systems and alternatively with help of linear and nonlinear MDOF systems. Parameters of the linear examples are elements of stiffness-, massand damping matrices or modal parameters alternatively. By nonlinear models the linear stiffness parameters are replaced by parameters characterizing the bilinear hysteresis which describes the restoring forces. The MATLAB program was used for programming of model, analysis and optimization.