Multi-Fidelity Bayesian Neural Network for Uncertainty Quantification in Transonic Aerodynamic Loads
This provides aerospace engineers with a more accurate and robust method for predicting aerodynamic loads with uncertainty estimates, though it appears incremental as it builds on existing multi-fidelity and Bayesian neural network techniques.
The paper tackles uncertainty quantification in transonic aerodynamic loads by implementing a multi-fidelity Bayesian neural network that fuses data from different fidelity sources using transfer learning, demonstrating that it outperforms state-of-the-art Co-Kriging in overall accuracy and robustness on unseen data.
Multi-fidelity models are becoming more prevalent in engineering, particularly in aerospace, as they combine both the computational efficiency of low-fidelity models with the high accuracy of higher-fidelity simulations. Various state-of-the-art techniques exist for fusing data from different fidelity sources, including Co-Kriging and transfer learning in neural networks. This paper aims to implement a multi-fidelity Bayesian neural network model that applies transfer learning to fuse data generated by models at different fidelities. Bayesian neural networks use probability distributions over network weights, enabling them to provide predictions along with estimates of their confidence. This approach harnesses the predictive and data fusion capabilities of neural networks while also quantifying uncertainty. The results demonstrate that the multi-fidelity Bayesian model outperforms the state-of-the-art Co-Kriging in terms of overall accuracy and robustness on unseen data.