Sequential Decision Process for Tradespace Exploration by Bounding Probabilistic Decision Criteria Using Mean-Risk Analysis

Uncertainty is an integral part of decision making. While performing tradespace analysis multiple design alternatives need to be compared with respect to uncertain decision criteria in order to identify non-dominated design alternatives. However when the decision criteria is obtained from a computationally intensive numerical analysis or from an experimental analysis it might not be feasible to precisely derive distributions of the decision criteria for all design alternatives in the tradespace. In this study it is hypothesized that the availability of precise distributions of decision criteria for all design alternatives in the tradespace is not necessary and appropriate decisions can be made on the basis of imprecise distributions of decision criteria. Key contribution of this study is to investigate an approach using mean-risk analysis to sequentially evaluate a tradespace of design alternatives by bounding and sequentially reducing the imprecision in evaluation of experimental/numerical performance. A sequential decision process is presented where models of increasing fidelity are used to discriminate dominated design alternatives from the tradespace on the basis of imprecise distributions of decision criteria. Application of the framework is demonstrated on a multi-objective discrete choice problem of designing a two bar truss.