Multidisciplinary design optimization of complex engineering systems for cost assessment under uncertainty

First, research is performed for investigating the performance of a Multidisciplinary Design Optimization (MDO) algorithm by integrating a particle swarm optimization (PSO) solver in both the system and discipline levels. The PSO solver is developed based on theoretical information available from the literature and the MDO framework is based on the Target Cascading (TC) method. The integrated MDO/PSO algorithm is employed for analyzing a conceptual ship design problem from the literature. Next, performance models are developed and employed within a gradient-based MDO framework for conducting a conceptual submarine design analysis. Four discipline level performances---internal deck area, powering, maneuvering, and structural analysis---are optimized simultaneously. The four discipline level optimizations are driven by a system level optimization which minimizes the manufacturing cost while at the same time coordinates the exchange of information and the interaction among the discipline level optimizations. The results from this coordinated MDO capture the interaction among disciplines and demonstrate the value that the MDO solution offers in consolidating the results to a single design which improves the discipline level objective functions while at the same time produces the highest possible improvement at the system level. Thirdly, a general method for improving the fidelity of cost estimation in the design of complex engineering systems is proposed. In this method, principal component analysis (PCA) and an adaptive Kriging method are used to increase the level of sophistication and predictive capability of existing cost assessment methodologies. Finally, an evidence theory-based uncertainty estimation algorithm is created and integrated into the cost assessment model developed earlier in order to capture the uncertainty surrounding the relationship between the system design variables and cost. The general cost assessment under uncertainty model is utilized as the system-level driver in the representative MDO study of a conceptual submarine design. A set of several MDO analyses which highlight the effects of different levels of uncertainty is also performed. The theoretical developments and the results from all four interrelated areas of research are summarized and discussed.