抄録
A robust design optimization of the convertible roof module mechanism is presented under consideration of the production uncertainty that has influence on the kinematic behavior of the convertible roof module mechanism. Random design variables are the positions of joints of the mechanism with uncertainty. In general, as the design goals for those mechanism behaviors are represented as the deviations over the roof module movements, it is very difficult to evaluate the analytical design sensitivity because the deviation is defined by using the max-value and the min-value over the parameter interval. Thus, this study introduces a meta-model technique to avoid the difficulty. Also, the sample variances for the design goals are approximated from those meta-models. In addition, a sequential approximation optimization technique is used to solve a robust design problem for the convertible roof module mechanism. The robust design problem has 8 random design variables with uncertainty. The proposed approach requires only 87 evaluations until converged.
