ABSTRACT
Traditionally, uncertainties in loading conditions, material properties, geometry or structural contour conditions have been included in the design process through hypotheses based on experience or engineering criteria such as the use of safety factors. Using these hypotheses, a simplified model is obtained, based on the nominal values of the variables and design parameters. However, the optimal solutions that have been reached with this deterministic approach have an optimum behavior only under conditions close to those fixed in the optimization process, and can deteriorate too much when the conditions move away from those of the design. This variability also exists in decision-making problems, where the choice of a solution depends on the assessment of the decision-makers involved in the problem. However, each decision-maker has different point of views or preferences, and when all the perspectives are taken into account, the uncertainty emerges.
The robust design optimization (RDO) is a methodology that makes possible to obtain solutions that are less sensitive to variations in the initial parameters, such as the characteristics of a material, the load conditions, or the different point of view of decision-makers. This methodology allows obtaining robust solutions that minimize both the mean value and the variability of the objective function. In this work, two examples of how the RDO methodology can help in obtaining solutions that are not very sensitive to the initial conditions are shown: a decision-making problem with different point of views of different decision-makers, and a structural problem where the initial parameters are considered variables.
