Comparing optimization algorithms for parameter identification of sigmoid model for MR damper

This paper proposes a new hybrid optimization technique that merges a differential evolution algorithm with a local strategy using the Nelder–Mead algorithm or simplex search algorithm and in Matlab software package, referred to as (fminsearch). To examine the variation in parameter estimation erros resulting from different optimization techniques. For a numerical model to exhibit good agreement with experimental values, it should prevent any clearances in the system and achieve an improved fit for the parameters of the Bouc–Wen-modified dynamic model. The study includes an experimental design to control the excitation current, frequency, and piston displacement. In this study, the model employed is the numerically parameterized model implemented by Wang, which utilizes experimental dynamic behavior of a commercial magnetorheological damper and applies a method to fit symmetric and asymmetric sigmoid functions using experimental data. These optimization algorithms are used to identify the sixteen parameters of the modified Bouc–Wen model.