The present research propose optimization methods that are robust in the sense that they produce
solutions insensitive to changes in the input parameters: these optimization methods are known
as Robust Design Techniques. The need for Robust Design method appears in many contests:
during the preliminary design process, the exact value of some input parameters is not known,
consequently the aim is try to look for a solution as less dependent on the unknown input parameters
as possible. The concept of robust optimization is demonstrated by using a 2-D airfoil shape
optimization problem. It has been observed [2] that minimizing drag at a single design point has
unintended consequences at nearby off-design points. Hicks and Vanderplaats demonstrate that
a direct optimization approach that minimizes drag at one mach number (e.g., M=0,75) actually
increases drag at nearby Mach number (e.g., M=0,70). To avoid this consequence the airfoil drag
minimization problem could be faced by means of an inverse optimization approach but the final
result could be almost the same of the single point design [7]. The conclusion is that the latter
techniques produce solutions that perform well in correspondence of the design point but have
poor off-design characteristics
