Design Optimization Methodologies

Abstract

The process of optimization involves choosing the best solution from a pool of potential candidate solutions. This article provides a description of some classes of problems and the optimization methods that solve them. These problems include the deterministic single-objective problem, the deterministic multiobjective problem, and the nondeterministic, stochastic optimization problem. The article presents several complementary approaches to solve a wide variety of single-objective and multiobjective mechanical engineering applications. Multiobjective optimization study and stochastic optimization studies are also discussed.

Optimal design of aeroacoustic airfoils with owl-inspired trailing-edge serrations

“…Related experiments [33] have shown that the designed CAS airfoils, which developed by University of Chinese Academy of Sciences, have stable aerodynamic performance. In this work, the new aerodynamic modules for serrated airfoils are introduced in section 2.3 and an optimal design framework for serrated airfoils is developed using Isight software [34]. The framework handles a set of design variables related to the parameterization of the serrated airfoil shape and solves the aerodynamic efficiency problem by taking into account the stringent aerodynamic and structural performance constraints defined in section 2.2, as schematically shown in figure 4.…”

An efficient application of Bayesian optimization to an industrial MDO framework for aircraft design

“…The CMDO framework is implemented as a workflow in Isight [4]. Isight is a process-integration, design-optimization software developed and distributed by Dassault Systemes.…”

“…In this section, we introduce six optimizers to solve the following constrained problem: Volume availability factor B[0] g [3] Landing gear spacing B[3] g [1] C L ma x retracted B [1] g [4] Wingtip twist deformation B[4] g [2] Cruise deck angle B [2] where f : R d → R is the objective function, g : R d → R m gives the inequality constraints which are expensive to evaluate and x ∈ Ω ⊂ R d is the vector of design variables. Two of the optimizers are included in the Isight software, which is mandatory to use because of CMDO and PMDO frameworks.…”

“…The Isight software [4] is delivered with many ready-to-use constrained optimizers. In this paper, we focus on two of the popular derivative-free Isight constrained optimizers: the evolutionary optimization algorithm (Evol) [26] and a variant of the Pointerdog algorithm (Pointer-2) [4].…”

“…The Pointer-2 strategy is an optimization framework based on an Isight proprietary algorithm that managed a set of well-known optimizers: Evol [26], the Hooke-Jeeves direct search method [27], the non-linear programming by quadratic Lagrangian (NLPQL) algorithm [28], the downhill simplex algorithm [29], the multi-function optimization system tool (MOST) technique [4] and the multi-objective particle swarm optimization (MOPSO) algorithm [30]. The choice of the set of optimizers used in the optimization process is made with respect to a classification system using the information available on the problem (i.e.…”