A systematic dataset generation technique applied to data-driven automotive aerodynamics
This work addresses a data generation bottleneck for aerodynamic shape optimization in the automotive industry, representing an incremental improvement in dataset creation methods.
The paper tackles the challenge of generating large and diverse training datasets for drag prediction in automotive aerodynamics by developing a novel strategy that systematically interpolates from a small number of starting points to create arbitrary samples. It demonstrates that convolutional neural networks achieve high performance in predicting drag coefficients and surface pressures, with promising results in extrapolation tests.
A novel strategy for generating datasets is developed within the context of drag prediction for automotive geometries using neural networks. A primary challenge in this space is constructing a training databse of sufficient size and diversity. Our method relies on a small number of starting data points, and provides a recipe to interpolate systematically between them, generating an arbitrary number of samples at the desired quality. We test this strategy using a realistic automotive geometry, and demonstrate that convolutional neural networks perform exceedingly well at predicting drag coefficients and surface pressures. Promising results are obtained in testing extrapolation performance. Our method can be applied to other problems of aerodynamic shape optimization.