Abstract
This paper studies the construction of an equivalent stiffness model for viscoelastic materials by genetic algorithm (GA). For this purpose the fractional derivative model of the stress-strain relationship are assumed for viscoelastic materials such as rubbers, and the parameters of the model are identified for the experimental frequency characteristics of the material. In minimizing the sum of errors between experimental values and the estimated values by the assumed model, a simple genetic algorithm procedure is employed in the present study. Finally the multicrestedness at this identification problem is studied by comparing the numerical results by using genetic algorithm and by using a conventional method.
