A new paradigm for concept selection in engineering design using multiobjective optimization

This thesis presents a new concept selection framework for engineering design using multiobjective optimization. The new framework capitalizes on the immense power of computational optimization by using it to evaluate disparate concepts early in the design process—before design freedoms have been significantly reduced. It is here, in the early phases of the engineering design process, that the impact of decision-making is the greatest. Interestingly, it is also in the early phases of design that the least rigorous decision-making methodologies are typically used. The new concept selection framework, set forth in this thesis, capitalizes on the efficiency and effectiveness of optimization to rapidly compare numerous designs, and characterize the tradeoff properties within the multiobjective design space. As such, the new framework differs significantly from traditional (non-optimization based) concept selection approaches where, comparatively speaking, significant time is often spent evaluating only a few points in the design space.; The new concept selection framework is based on the important principle of Pareto optimality, which defines an important class of optimal solutions to multiobjective optimization problems. The complete set of Pareto solutions is known as the Pareto frontier. Pareto frontiers are particularly useful in engineering design because they characterize the tradeoffs between competing design objectives.; Under the developments of this thesis, disparate design concepts are evaluated using a so-called s-Pareto frontier; this frontier originates from the Pareto frontiers of various disparate concepts, and is the Pareto frontier for the set of concepts. Similar to other Pareto frontiers, the s-Pareto frontier can be used to characterize the tradeoffs between design objectives. However, unlike other Pareto frontiers, the s-Pareto frontier can be used to characterize the tradeoffs between disparate design concepts. This property of the s-Pareto frontier is what makes it extremely useful for decision-making in conceptual design. The significance of the s-Pareto frontier is that it makes it possible to use optimization to explore the design space in the early phases of design—as it pertains to more than one concept.