抄録
This paper proposes a mini-max type formulation for strict robust design optimization under correlative variation based on design variation hyper sphere and quadratic polynomial approximation. The proposed method aims that all points within distribution region strictly satisfy constraints and the maximum of objective function is minimized. For this purpose, the objective function and the constraints are approximated as quadratic polynomials over the region and the design space is diagonalized and isoparameterized into a hyper sphere. The maximum of each function is obtained by mathematical means with less computation cost based on these transformations within the inner loop of mini-max type optimization computation. The validity of the proposed formulation is ascertained through its application to a two-dimensional numerical example.
