Decision Topology Assessment in Engineering Design Under Uncertainity

摘要

The implications of decision analysis (DA) on engineering design are well known. Recently, the authors proposed decision topologies (DT) as a visual method for making design decisions and proved that they are consistent with normative decision analysis. This paper addresses the practical issue of assessing DTs for a decision maker (DM) using their responses, particularly under uncertainty. This is critical to encoding decision maker preferences so that further analysis and mathematical optimization can be performed using the correct set of preferences. We show how multiattribute DTs can be directly assessed from DM responses. Four methods are shown to evolutionarily assess DTs among which one that requires the DM to rank alternatives and another where a utility function is first assessed. It is also shown that preferences under uncertainty can be easily incorporated. In addition, we show that topologies can be constructed using single attribute topologies similarly to multi-linear functions in utility analysis. This incremental construction simplifies the process of topology construction. The reverse problem of inferring single attribute DTs is also presented. The proposed assessment methods are used on a design decision making problem of a welded beam.