An Efficient Evolutionary Multi-Objective Approach for Robust Design of Multi-Stage Space Launch Vehicle
Saqlain Akhtar, He Linshu · 11th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference · 2006
*† Many candidate concepts of Space Launch Vehicles (SLVs) have been proposed around the world by incorporating deterministic optimization approach. Our aim in this study is to apply evolutionary approach for robust conceptual design of SLV considering trajectory optimization. In this study we proposed a method for estimating the robustness of a solution by exploiting the information available in the current population of the evolutionary algorithm, with out any additional fitness evaluations. In this evolutionary robust optimization approach, the basic idea is to define a neighborhood of a solution and thus to estimate the local mean and variance of a solution. Thus, a trade-off between optimality and robustness can be realized with the help of evolutionary multi-objective optimization. Unlike previous methods in the literature, our proposed method does not use gradient information, and hence is applicable to multi-objective optimization problems that have nondifferentiable and/or discontinuous objective and/or constraint functions. The proposed method is implemented on a complex SLV conceptual design test problem. Monte Carlo simulation studies have been conducted for sensitivity analysis to verify the robustness of proposed method. Nomenclature _ i x σ = average standard deviation µf