Comparison of Optimization Strategies for High-Lift Design

Abstract

The design of high-lift systems represents a challenging task within the aerospace community, being a multidisciplinary, multi-objective and multi-point problem. The DeSiReH (Design, Simulation and Flight Reynolds Number Testing for Advanced High-Lift Solution) project, funded by European Commission under the 7th Framework Program, aimed at improving the aerodynamics of high-lift systems by developing, in a coordinated approach, both efficient numerical design strategies and measurement techniques for cryogenic conditions. Within DeSiReH, different partners used several numerical automatic optimization strategies for high-lift system design purposes. A realistic multi-objective and multi-point optimization problem was defined and solved by adopting different flow models dimensionality, meshing techniques, geometry parameterization and optimization strategies. Special attention was devoted to perform a fair comparison of the results and useful information were obtained about trends, pros and cons of the approaches used. The outcome of these activities is that an efficient design

High-Lift Devices Topology Optimisation using Structured-Chromosome Genetic Algorithm

“…Therefore, the higher the tournament size, the higher the pressure to select only the high-fitness candidates, with the possible drawback of letting the population collapse toward a local optimum. The values tested for the population size are [5,10,20] and for the tournament size are [2,4]. However, the combination with population size equal to 5 and tournament size equal to 4 cannot be run because of the implementation of the tournament selection in SCGA.…”

“…The reference run is made using the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) [26] library has been already tested for a bunch of aerodynamic design optimisation problems facing multi-element airfoils [5], [6]. It is a stochastic optimisation algorithm based on selfadaptation of the covariance matrix of a multivariate normal distribution.…”

“…As discussed in [5], the HLD design is an arduous task because it is as a multidisciplinary (aerodynamics, structures, systems integration disciplines are involved), multi-objective (generally in conflict) [6], and multi-point (designed for takeoff, climbing, approach, landing flight phases) [7] design problem. Thus, HLD design is very demanding from a computational point of view.…”

“…Classically, the design of HLDs is a two-stage procedure, first a promising configuration type (number of airfoil elements) is selected, and second, the position and shape of the elements are obtained through optimisation, often employing heuristic algorithms [5]. Furthermore, because most of them are limited to continuous numeric variables and fixed-length search spaces, optimisation algorithms are assigned to solve a circumscribed part of the original design problem exclusively.…”

Sustainable Development Analysis Model Using Multi-objective Optimization and Regional Efficiency Evaluation

“…e clashes between social and economic progress, resource conservation, and environmental protection are becoming more apparent. Issues including resource scarcity, pollution, and ecological imbalance have exposed the faults of the classic development model [7]. Despite China's great achievements in economic development, the contradiction of an unreasonable economic structure has built up over time, costing the country dearly in terms of demand imbalances, uncoordinated supply structures, low factor utilisation efficiency, large environmental damage, and unreasonable spatial layout, among other things [8].…”