Optimization based on parameter moments estimation for robust design

This paper deals with robust design based on analytical models. We consider that variability on design parameters are represented by the parameter Mean and Standard Deviation. We propose in this article a new robust design approach that allows the designer either to find a robust solution in a reasonable execution time or to be informed of the non-existence of a robust solution for his design problem. The process combines a reformulation of the model to integrate parameter uncertainties and a deterministic optimization algorithm with new specifications based on robust objectives. The engineering application of an electrical actuator design is presented and used to show the implementation and the effectiveness of this new robust approach.