The University of Tokyo · Engineering
Professor Kazuo Yonekura's research lab specializes in the integration of machine learning and physics-based modeling for engineering design and prediction, with a strong focus on aerodynamics, shape optimization, and uncertainty quantification. The lab develops advanced deep learning techniques—such as conditional GANs, VAE-WGAN hybrids, and physics-guided neural networks—to generate high-performance airfoil shapes and improve the reliability of short-term weather forecasts. A key research direction involves enhancing the physical consistency and interpretability of AI models through methods like Monte Carlo dropout for uncertainty estimation and guided training that embeds physical constraints without requiring gradient computation. The lab also emphasizes practical applications in energy management, transportation, and industrial design by leveraging data-driven approaches with robustness and generalization.
Figures are computed from collected data and may differ slightly.
This paper proposes a very-short-term, i.e., less than 1-hour, local weather forecast method. In general, a short-term weather forecast within 3 hours is difficult due to lack of surface weather data and limitations of computation resources. However, such a short-term prediction is getting more and more anticipated in several industrial situations such as transportation, retailing business, agriculture, and energy management as well as our daily life. To keep up with this huge demands, services
This paper proposes a shape parameterization method using a principal component analysis (PCA) for shape optimization. The proposed method is used as a preprocessing tool of parametric optimization algorithms, such as genetic algorithms (GAs) or response surface methods (RSMs). When these parametric optimization algorithms are used, the number of parameters should be small while the design space represented by the parameters should be able to represent a variety of shapes. In order to define the
A machine learning method was applied to solve an inverse airfoil design problem. A conditional VAE-WGAN-gp model, which couples the conditional variational autoencoder (VAE) and Wasserstein generative adversarial network with gradient penalty (WGAN-gp), is proposed for an airfoil generation method, and then, it is compared with the WGAN-gp and VAE models. The VAEGAN model couples the VAE and GAN models, which enables feature extraction in the GAN models. In airfoil generation tasks, to generate
Open papers in the app to read, cite, and organize with AI.
