Corpus ID: 237941054

Deep Learning for Multi-Fidelity Aerodynamic Distribution Modeling from Experimental and Simulation Data

  title={Deep Learning for Multi-Fidelity Aerodynamic Distribution Modeling from Experimental and Simulation Data},
  author={Kai Li and Jiaqing Kou and Weiwei Zhang},
  • Kai Li, J. Kou, Weiwei Zhang
  • Published 27 September 2021
  • Physics
The wind-tunnel experiment plays a critical role in the design and development phases of modern aircraft, which is limited by prohibitive cost. In contrast, numerical simulation, as an important alternative paradigm, mimics complex flow behaviors but is less accurate compared to experiment. This leads to the recent development and emerging interest in applying data fusion for aerodynamic prediction. In particular, the accurate prediction of aerodynamic with lower computational cost can be… Expand


Robust Design of a Reentry Unmanned Space Vehicle by Multifidelity Evolution Control
This paper addresses the preliminary robust design of a small/medium–scale reentry unmanned space vehicle. A hybrid optimization technique is proposed that couples an evolutionary multi-objective a...
Data-Driven Aerospace Engineering: Reframing the Industry with Machine Learning
This review will explore the opportunities and challenges of integrating data-driven science and engineering into the aerospace industry, and focus on the critical need for interpretable, generalizeable, explainable, and certifiable machine learning techniques for safety-critical applications. Expand
Data-driven modeling for unsteady aerodynamics and aeroelasticity
  • J. Kou, Weiwei Zhang
  • Computer Science
  • Progress in Aerospace Sciences
  • 2021
Three typical data-driven aerodynamic methods are introduced, including system identification, feature extraction and data fusion, which help to gain physical insights on flow mechanism and have shown great potential in engineering applications like flow control, aeroelasticity and optimization. Expand
Multi-fidelity Bayesian Neural Networks: Algorithms and Applications
The results demonstrate that the present method can capture both linear and nonlinear correlation between the low- and high-fidelity data adaptively, identify unknown parameters in PDEs, and quantify uncertainties in predictions, given a few scattered noisy high- fidelity data. Expand
Physics-informed machine learning
| Despite great progress in simulating multiphysics problems using the numerical discretization of partial differential equations (PDEs), one still cannot seamlessly incorporate noisy data into… Expand
Research on refined reconstruction method of airfoil pressure based on compressed sensing
A refined reconstruction method of airfoil surface pressure based on compressed sensing, which can reconstruct the pressure distribution with high precision with less pressure measurement data is proposed. Expand
Turbulence closure for high Reynolds number airfoil flows by deep neural networks
This work constructs black-box algebraic models to substitute the traditional turbulence model by the artificial neural networks (ANN) rather than correcting the existing turbulence models in most of current studies, and shows the prospect of turbulence modeling by machine learning methods. Expand
A composite neural network that learns from multi-fidelity data: Application to function approximation and inverse PDE problems
The proposed composite neural network (NN) is capable of learning both the linear and complex nonlinear correlations between the low- and high-fidelity data adaptively and can be readily extended to very high-dimensional regression and classification problems involving multi-f fidelity data. Expand
Development of aMulti-Fidelity Reduced-OrderModel Based onManifold Alignment
  • AIAA AVIATION 2020 FORUM, 2020, p. 3124.
  • 2020