Fast reconstruction of aerodynamic shapes using evolutionary algorithms and virtual nash strategies in a CFD design environment

Abstract

a b s t r a c tThis paper compares the performances of two different optimisation techniques for solving inverse problems; the first one deals with the Hierarchical Asynchronous Parallel Evolutionary Algorithms software (HAPEA) and the second is implemented with a game strategy named Nash-EA. The HAPEA software is based on a hierarchical topology and asynchronous parallel computation. The Nash-EA methodology is introduced as a distributed virtual game and consists of splitting the wing design variables -aerofoil sections -supervised by players optimising their own strategy. The HAPEA and Nash-EA software methodologies are applied to a single objective aerodynamic ONERA M6 wing reconstruction. Numerical results from the two approaches are compared in terms of the quality of model and computational expense and demonstrate the superiority of the distributed Nash-EA methodology in a parallel environment for a similar design quality.Crown

Double Shock Control Bump design optimization using hybridised evolutionary algorithms

“…Each node (Node0 ~ Node6) belonging to the different hierarchical layer can be handled by a different EA code. The second method hybridises HAPMOEA by applying a concept of Nash-Equilibrium instead of the concept of hierarchical multi-population topology [4,17]. The Nash-Game players choose their own strategy to improve their own objective.…”

“…In addition, Lee et al [3] hybridised NSGA-II with Nash-Game strategy to study a role of Nash-Players in Hybrid-Game by solving multi-objective mathematical design problems convex, discontinuous, etc. And Hierarchical Asynchronous Parallel Multi-Objective Evolutionary Algorithm (HAPMOEA) [4] is also hybridised to solve a real-world robust multidisciplinary design problem. Numerical results show that the Hybrid-Game improves 70% of HAPMOEA performance while producing better Pareto optimal solutions.…”

Multiobjective Design Optimization Using Nash Games

“…In this context, the splitting of territory becomes a domain decomposition which could accelerate the convergence and help avoiding local optima. This approach was proposed by Périaux (47). As illustration, we consider a wing design problem, which requires the definition of global geometrical characteristics, such as span, root/tip length ratio, angle of attack, twist angle, sweep angle, and so on, as well as local geometrical features that determine the wing section.…”

Double-shock control bump design optimization using hybridized evolutionary algorithms

“…In addition, Lee et al [3] hybridized NSGA-II with Nash-Game strategy to study a role of Nash-Players in Hybrid-Game by solving multi-objective mathematical test cases; non-uniformly distributed non-convex, discontinuous, and mechanical design problem. Their research also shows that hierarchical asynchronous parallel multi-objective evolutionary algorithm (HAPMOEA) [4] can also be hybridized to solve a real-world robust multidisciplinary design problem. Numerical results show that the Hybrid-Game improves 70 per cent of HAPMOEA performance while producing better Pareto optimal solutions.…”