Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 2 of 2
Full-Text Articles in Aerospace Engineering
A Convolutional Neural Network Model Based On Multiscale Structural Similarity For The Prediction Of Flow Fields, Yifu An, Xiaosong Du, Joaquim R.R.A. Martins
A Convolutional Neural Network Model Based On Multiscale Structural Similarity For The Prediction Of Flow Fields, Yifu An, Xiaosong Du, Joaquim R.R.A. Martins
Mechanical and Aerospace Engineering Faculty Research & Creative Works
We have seen the emerging applications of deep neural networks for flow field predictions in the past few years. Most of the efforts rely on the increased complexity of the model itself or take advantage of novel network architectures, such as convolutional neural networks (CNN). However, reaching low prediction error cannot guarantee the quality of the predicted flow fields in terms of the perceived visual quality. This work introduces the multi-scale structural similarity (MS-SSIM) index method for flow field prediction. First, we train CNN models using the commonly used root mean squared error (RMSE) loss function as the reference. Then …
Novel Adaptive Sampling Algorithm For Pod-Based Non-Intrusive Reduced Order Model, Jiachen Wang, Xiaosong Du, Joaquim R.R.A. Martins
Novel Adaptive Sampling Algorithm For Pod-Based Non-Intrusive Reduced Order Model, Jiachen Wang, Xiaosong Du, Joaquim R.R.A. Martins
Mechanical and Aerospace Engineering Faculty Research & Creative Works
The proper orthogonal decomposition (POD) based reduced-order model (ROM) has been an effective tool for flow field prediction in the engineering industry. The sample selection in the design space for POD basis construction affects the ROM performance sensitively. Adaptive sampling can significantly reduce the number of samples to achieve the required model accuracy. In this work, we propose a novel adaptive sampling algorithm, called conjunction sampling strategy, which is based on proven strategies. The conjunction sampling strategy is demonstrated on airfoil flow field prediction within the transonic regime. We demonstrate the performance of the proposed strategy by running 10 trials …