Applications of surrogate-assisted and multi-fidelity multi-objective optimization algorithms to simulation-based aerodynamic design

Abstract

Purpose The purpose of this work is to apply and compare surrogate-assisted and multi-fidelity, multi-objective optimization (MOO) algorithms to simulation-based aerodynamic design exploration. Design/methodology/approach The three algorithms for multi-objective aerodynamic optimization compared in this work are the combination of evolutionary algorithms, design space reduction and surrogate models, the multi-fidelity point-by-point Pareto set identification and the multi-fidelity sequential domain patching (SDP) Pareto set identification. The algorithms are applied to three cases, namely, an analytical test case, the design of transonic airfoil shapes and the design of subsonic wing shapes, and are evaluated based on the resulting best possible trade-offs and the computational overhead. Findings The results show that all three algorithms yield comparable best possible trade-offs for all the test cases. For the aerodynamic test cases, the multi-fidelity Pareto set identification algorithms outperform the surrogate-assisted evolutionary algorithm by up to 50 per cent in terms of cost. Furthermore, the point-by-point algorithm is around 27 per cent more efficient than the SDP algorithm. Originality/value The novelty of this work includes the first applications of the SDP algorithm to multi-fidelity aerodynamic design exploration, the first comparison of these multi-fidelity MOO algorithms and new results of a complex simulation-based multi-objective aerodynamic design of subsonic wing shapes involving two conflicting criteria, several nonlinear constraints and over ten design variables.

Multi-fidelity Multi-Objective Optimization of a High-Altitude Propeller

“…In this class of methodologies, it is often that a Pareto front approximation is found with the help of a standard multi-objective optimizer (like the NSGA-II or MOPSO) while the objectives are modeled by multi-fidelity metamodels. Then different criteria are used to better resolve the Pareto front approximation and choose the solver fidelity for each iteration [38][39][40]. Similarly, Kontogiannis et al [41] used the Expected Improvement (EI) of each objective to find a non-dominated front of EIs with the help of a population-based multi-objective algorithm.…”

A hybrid computational method for optimization design of bistable compliant mechanism

“…On the contrary, FEM, which is a numerical method, is the best for compliant mechanism modeling. More recently, surrogate-based approaches, computational intelligence and machine learning are known as powerful approaches to model complex systems (Yüksel and Sezgin, 2010; Jiang et al , 2015; Guo et al , 2016; Amrit and Leifsson, 2019).…”

“…Over the past decades, optimization techniques have been proposed, such as GA (Keshtiara et al , 2019), particle swarm optimization (PSO) (Chau et al , 2019), differential evolution (Dao et al , 2017a), cuckoo search (Dao et al , 2017b), INSGA-II (Bu et al , 2019), ANN-based GA (Soepangkat et al , 2019), surrogate-assisted MOO algorithms (Amrit and Leifsson, 2019), improved plant growth simulation algorithm–PSO hybrid algorithm (Jiang et al , 2020) and other algorithms (Chernogorov et al , 2017; Senkerik et al , 2017; Zatloukal and Znoj, 2017; Dinh-Cong et al , 2018). Presently, algorithms with free parameters were developed, e.g.…”