Extension of Analytical Target Cascading Using Augmented Lagrangian Coordination for Multidisciplinary Design Optimization

Abstract

Analytical target cascading (ATC) is a method originally developed for translating system-level design targets to design specifications for the elements comprising the system. ATC has also been shown to be useful for coordinating distributed design optimization of hierarchical, multilevel systems. The traditional ATC formulation uses a hierarchically decomposed problem structure, in which coordination is performed by communicating target and response values between parents and children. This paper presents two extensions of the ATC formulation to allow non-hierarchical target-response coupling between subproblems and to introduce system-wide constraints that depend on local variables of two or more subproblems. The ATC formulation with these extensions belongs to a subclass of augmented Lagrangian coordination, and has thus converge properties under the usual convexity and continuity assumptions. A supersonic business jet design problem reported earlier in the literature is used to illustrate these extensions.