Constrained exploration of trade spaces

Use of integrated multi-disciplinary models opens up the conceptual design space substantially. In order to take advantage of this opportunity to improve design quality, we have developed an approach that integrates a constraint management system (CMS) to define performance and cost relationships, a sampling methodology to define a trade space, and a trade-off capability to rapidly explore this space. The sampling methodology allows a mixture of spanning, factorial, and probabilistic operators. Further, these operators are constrained by the relationships within the CMS to limit exploration to feasible portions of the trade space. The trade-off capability uses multiple-objective optimization techniques to obtain Pareto-optimal designs with respect to competing objectives within this constrained trade space. In this paper, we describe the trade space exploration capability and how it has been used for evaluation of alternative architectures, for example in the design of space operations vehicle, using integrated cost-performance trade studies.