427 / 2022-08-01 17:18:50

Parameter optimization of a seven-degree-of-freedom seated human model using a hybrid optimization method

human model,parameter optimization,gradient-based algorithm,Genetic Algorithm (GA),hybrid optimization methodology

This study focuses on the development of a seven degree-of-freedom (DOF) seated human model and proposes a hybrid optimization method to estimate the parameters of the human model. Human skin and vehicle seats have viscoelastic properties, so their mechanical properties vary and are difficult to measure. To estimate effective stiffness and damping coefficients of the human skin and vehicle seats, we designed a constrained minimization optimization problem that involves fitting the human model to experimental data obtained from a whole-body vibration experiment. First, we explored the capabilities of gradient-based and genetic algorithms as solution search strategies for the human parameters. Then, we applied the proposed hybrid optimization method, which combines the advantages of the two previous algorithms in some systematically designed steps. The hybrid optimization method not only yielded optimization parameters that logically match the human model design in terms of numerical stability, but it also provided superior computational performance with the least error and the fastest optimization speed. Therefore, the hybrid optimization method enhanced the robustness of the optimization process. Finally, the frequency softening phenomenon was observed in the dynamic response of the model as the excitation magnitude increased from 0.5 to 2.0 m/s² rms.