Robust design optimization by spline dimensional decomposition

This article highlights new spline-empowered computational methods for solving robust design optimization (RDO) problems in complex mechanical systems. The are predicated on a spline dimensional decomposition (SDD) of high-dimensional, discontinuous, or nonsmooth stochastic response statistical moment analysis, novel fusion SDD and score functions calculating the second-moment sensitivities with respect to variables, standard gradient-based algorithms. New closed-form formulae derived that concurrently determined along moments. depend how sensitivity analyses engaged an algorithm, engendering direct multi-point single-step processes. Numerical results reveal proposed deliver accurate computationally efficient optimal solutions RDO problems, including industrial-scale shape robotic gripper jaw.