A hybrid approach based on evolutionary strategies and interval arithmetic to perform robust designs

Abstract

This chapter proposes an approach based on the use of Evolutionary Algorithms (EA) and Interval Arithmetic (IA) as an alternative technique to obtain a robust system design. EA are heuristics to optimise a given function, while IA is used as a checking technique to guarantee the feasibility of the design. IA is able to consider simultaneously the effects of uncertainty of all of the parameters on a performance function providing strict bounds (minimum and maximum values) with only one evaluation. EA and IA are used to obtain, by an iterative process, a robust design, that is, the maximum size of each variable deviation that allows complying with a set of specifications. The proposed approach is an indirect method based on optimization instead of a direct method based on mapping from the output into the input space. Numerical examples from the reliability field illustrate the application of the approach using single and multiple objective formulations.