Abstract:<p>Adjoint-based optimization methods are attractive for aerodynamic shape design primarily due to their computational costs being independent of the dimensionality of the input space and their ability to generate high-fidelity gradients that can then be used in a gradient-based optimizer. This makes them very well suited for high-fidelity simulation based aerodynamic shape optimization of highly parametrized geometries such as aircraft wings. However, the development of adjoint-based solvers involve careful mathematical treatment and their implementation require detailed software development. Furthermore, they can become prohibitively expensive when multiple optimization problems are being solved, each requiring multiple restarts to circumvent local optima. In this work, we propose a machine learning enabled, surrogate-based framework that replaces the expensive adjoint solver, without compromising on predicting predictive accuracy. Specifically, we first train a deep neural network (DNN) from training data generated from evaluating the high-fidelity simulation model on a model-agnostic design of experiments on the geometry shape parameters. The optimum shape may then be computed by using a gradient-based optimizer coupled with the trained DNN. Subsequently, we also perform a gradient-free Bayesian optimization, where the trained DNN is used as the prior mean. We observe that the latter framework (DNN-BO) improves upon the DNN-only based optimization strategy for the same computational cost. Overall, this framework predicts the true optimum with very high accuracy, while requiring far fewer high-fidelity function calls compared to the adjoint-based method. Furthermore, we show that multiple optimization problems can be solved with the same machine learning model with high accuracy, to amortize the offline costs associated with constructing our models. Our methodology finds applications in the early stages of aerospace design.</p>

Artificial neural network based inverse design: Airfoils and wings

Aerodynamic Optimization Design of Airfoil Configurations Based on Cascade Feedforward Neural Network

Database Self-Expansion Based on Artificial Neural Network: an Approach in Aircraft Design

Inverse Design of Airfoils Using Convolutional Neural Network and Deep Neural Network

Inverse design optimization framework via a two-step deep learning approach: application to a wind turbine airfoil

An Iterative Inverse Design Method of Turbomachinery Blades by Using Proper Orthogonal Decomposition

Aerodynamic Design Optimization by Using a Continuous Adjoint Method

A novel approach to inverse design of wind turbine airfoils using tandem neural networks

Intelligent Aerodynamic Design for Airfoil Based on Artificial Neural Network Method

Neural network design for data-driven prediction of target geometry for an aerodynamic inverse design algorithm

Artificial neural network for preliminary design and optimisation of civil aero-engine nacelles

Wind turbine blade design with airfoil shape control using invertible neural networks

A review of the artificial neural network surrogate modeling in aerodynamic design

An inverse design method for supercritical airfoil based on conditional generative models

Framework of nacelle inverse design method based on improved generative adversarial networks

Enhanced data efficiency using deep neural networks and Gaussian processes for aerodynamic design optimization

Inverse design optimization of the compressor cascade airfoil assisted by active subspace approach

Generative Inverse Design of Aerodynamic Shapes Using Conditional Invertible Neural Networks

New approach of inverse design of transonic compressor rotor blade via prescribed isentropic Mach distributions without modification of governing equations

Efficient Inverse Design Optimization through Multi-fidelity Simulations, Machine Learning, and Search Space Reduction Strategies

Airfoil shape optimization using genetic algorithm coupled deep neural networks