Generative Design of Ship Propellers using Conditional Flow Matching
This addresses the need for versatile and efficient design generation in engineering, specifically for ship propellers, though it is incremental as it builds on existing generative AI methods applied to a new domain.
The paper tackles the problem of generating ship propeller designs that meet specified performance targets by using conditional flow matching to create a bidirectional mapping between design parameters and noise conditioned on performance, enabling multiple valid designs for the same targets, with examples showing distinct geometries achieving nearly identical performance.
In this paper, we explore the use of generative artificial intelligence (GenAI) for ship propeller design. While traditional forward machine learning models predict the performance of mechanical components based on given design parameters, GenAI models aim to generate designs that achieve specified performance targets. In particular, we employ conditional flow matching to establish a bidirectional mapping between design parameters and simulated noise that is conditioned on performance labels. This approach enables the generation of multiple valid designs corresponding to the same performance targets by sampling over the noise vector. To support model training, we generate data using a vortex lattice method for numerical simulation and analyze the trade-off between model accuracy and the amount of available data. We further propose data augmentation using pseudo-labels derived from less data-intensive forward surrogate models, which can often improve overall model performance. Finally, we present examples of distinct propeller geometries that exhibit nearly identical performance characteristics, illustrating the versatility and potential of GenAI in engineering design.