A novel generative approach to the parametric design and multi-objective optimization of horizontal axis tidal turbines.

This article focuses on the development and application of a variational autoencoder generative adversarial network (VAEGAN) model for the design and multi-objective optimization of horizontal axis tidal turbine (HATT) blades. The VAEGAN model compresses complex three-dimensional blade geometric features into a low-dimensional latent space, enabling efficient generation of diverse, smooth, and physically interpretable blade designs. Integrated with blade element momentum theory and finite element analysis, the model facilitates simultaneous optimization of hydrodynamic performance and structural strength, outperforming traditional parametric methods in convergence speed and solution quality. The study demonstrates the potential of deep generative models to expand design spaces and reduce computational costs in industrial turbine blade design, while suggesting future work on latent space interactions and inclusion of material and structural parameters for comprehensive optimization.