Success in engineering design is greatly dependent on the decisions being made by the designers throughout a design process. In addition, engineering problems are mostly complex in nature, which leaves the designer performing complex trade-off decisions in order to pick a point, or set of points with which to proceed. In order to increase the level of confidence in the decision making process, it is necessary to accurately translate the designer's preferences as well as incorporate the inherent and inevitable presence of uncertainty within the design process.;This thesis proposes a new method of incorporating the designer's preferences based on the concept of Physical Programming [7] preference ranges as well as representing uncertainty within a given multi-objective optimization problem (MOP) using a visualization tool method named Hyper Radial Visualization (HRV). By defining the preferences and uncertainty present in a MOP, the method will differentiate the points as well as sort out the infeasible ones based on the uncertainty defined by the designer.;Two case studies are presented in this thesis to illustrate the mechanism behind the method as well as the impact the method has on different problems with varying levels of complexity. The first case study illustrates the mechanism driving the method for a real engineering problem. The second case study illustrates the method in a high dimensional problem. Finally, based on the results obtained in the case studies, future work is addressed.
展开▼
