A Comparison on the Performance of Heuristic Methods of Search for Parameter Characterization of the Bouc/Wen Model of Mechanical Hysteresis

The Bouc/Wen model of mechanical hysteresis may be stated as a system of coupled parametric nonlinear differential equations of third order. This model is based on the equation of motion with damping for a single degree-of-freedom (DOF) system, which is able to reproduce the dynamics of a wide variety of hysteresis shapes, depending on the values of the seven parameters present in the equations. The characterization of the parameters that describe the specific shapes of hysteresis loops is not an easy task, due to the complexity of the model and the great amount of local minima in the search space. Accordingly, heuristic methods of search (optimization) must be carried out in order to curve-fit the model to the trial data sets, in order to extract damping information. Previous works on this topic deal with optimization using Genetic Algorithms (GA). This paper presents a new approach, using Particle Swarm Optimization (PSO). The performances of both methods are then compared when used for artificial and experimental hysteresis loops obtained for sandwich composite materials under test. Some damping measurement techniques often used by researchers are also explained.