| This thesis examines decomposition based procedures in the optimal design of large-scale multidisciplinary systems. The use of formal optimization methods in such systems is complicated by the presence of a large number of design variables and constraints. Decomposition reduces a large-scale system into a sequence of smaller, more tractable subsystems, each with a smaller set of design variables and constraints. The decomposed subsystems, however, are not totally decoupled, and design changes in one subsystem may have a profound influence on changes in other subsystems. The present work examines the effectiveness of counterpropagation (CP) neural networks as a tool to account for this coupling. This capability derives from a pattern completion capability of such networks. The proposed approach is implemented for a class of structural design problems where the decomposed subsystems exhibit hierarchy, i.e., there is a distinct chain of command in the nature of couplings between the subsystems. The role of artificial neural networks is also explored in the context of concurrent subspace optimization (CSSO) where this decomposition based approach is applicable to problems where no distinct hierarchy of influences can be identified.;Essential components of decomposition based design methods are strategies to identify a topology for problem decomposition, and to develop coordination strategies which account for couplings among the decomposed problems. The present thesis examines the effectiveness of artificial neural networks as a tool to both account for the coupling, and to develop methods to coordinate the solution in the different subproblems. The solution process for decomposition based design is further enhanced by a novel approach of using Intelligent Agents (IA's). This agent-based paradigm provides the necessary support structure for representing salient characteristics of the design, and for coordinating the solutions in different subproblems. The CSSO method is adapted in this environment, and results obtained with representative application problems indicate that it permits multidisciplinary design teams to work effectively on different parts of the problem by automating individual tasks, information sharing, and facilitating coordination among the engineering teams. |
