Design Optimization of Hierarchically Decomposed Multilevel Systems under Uncertainty

This paper presents a methodology for design optimization of decomposed systems in the presence of uncertainties. We extend the analytical target cascading (ATC) formulation to probabilistic design by treating stochastic quantities as random variables and parameters and posing reliability-based design constraints. We model the propagation of uncertainty throughout the multilevel hierarchy of elements that comprise the decomposed system by using the advanced mean value (AMV) method to generate the required probability distributions of nonlinear responses. We utilize appropriate metamodeling techniques for simulation-based design problems. A simple yet illustrative hierarchical bi-level engine design problem is used to demonstrate the proposed methodology.