Pacheco, J.E., Amon, C.H., Finger, S. "Using Bayesian Models in Preliminary Design", International Design Conference - Design 2002, Dubrovnik, 2002.

This work is developing analytical methods that will enable designers to create models with different levels of detail and accuracy depending on the current state of the design. During preliminary design, many parameters are not known with great precision. Additionally, the design requirements are often vague and change frequently, so single point solutions are not the best approach. Instead, we propose to create models that allow for the incorporation of new information as the design progresses [Osio and Amon 1996]. A major advantage of these models is that they can be continuously refined. Thus, an approximate model developed for preliminary design can evolve throughout the design process. By the time the final design stage is reached, the model can be highly accurate in the reduced region of the design space close to the optima. We believe that having a methodology that allows designers to create models at any desired level of accuracy will improve the selection of initial ideas at the conceptual design stage, decrease design cycle time, and reduce costs in optimisation.

Our methodology creates surrogate models using a Bayesian framework. The surrogate is an analytical model of the expected response for a set of design parameters. This methodology does not require the assumption of any specific form of the response; instead, models are defined in terms of the correlation between sampling sites, assuming the response is a realization of a stochastic process. We do assume that the designer is able to evaluate the system response for a set of design parameters using analysis, computation models, or physical experiments.