A NON-DETERMINISTIC APPROACH TO CONCEPT SELECTION USING S-PARETO FRONTIERS

摘要

The most significant design decisions are typically made during the conceptual phase of the engineering design process, when critical design features are proposed, evaluated and selected. In this paper, we explore the critical task of concept selection and propose a non-deterministic, optimization-based approach for selecting the most promising concept. The method presented in this paper builds upon the recently-proposed s-Pareto based concept selection approach. Within the framework of the s-Pareto approach, so-called s-Pareto frontiers are obtained by using the definition of Pareto optimality to identify Pareto optimal solutions that pertain to a set of distinct concepts. These s-Pareto frontiers are used to assess the tradeoffs between various proposed concepts during conceptual design. The s-Pareto approach is a marked departure from traditional concept selection methods and from the traditional use of Pareto frontiers. In this work the s-Pareto approach is extended to include uncertainties caused by stochastic design parameters as well as low model fidelity. More specifically, the reliability of design decisions is accounted for in the decision-making process. Two approaches are presented for performing non-deterministic concept selection. Two examples are given that support the approach.