Parameter identification of rheological models using optimization algorithms

Abstract

The paper presents results of the application of optimization software GAMS and CONOPT solver for nonlinear optimization problems to identify multi-parameter elastomer rheological models. The solution is based on the experimentally set frequency-dependent dynamic stiffness of the elastomeric parts. Measured waveforms of dynamic stiffness are approximated using multi-parameter rheological models. The CONOPT solver has shown a significantly faster convergence and better accuracy of the calculation compared with the classical non-linear Gauss-Newton least square optimization technique. © Springer International Publishing Switzerland 2014.