arXiv:2607.06999v1 Announce Type: new Abstract: This paper presents a physics-guided machine learning (PGML) framework for fuel density prediction, integrating physics constraints and domain knowledge into deep learning models to enhance model accuracy and stability. We explore three deep learning architectures -- ConvLSTM, Adaptive Fourier Neural Operator (AFNONet), and Video Vision Transformer (ViViT) -- to model the spatiotemporal evolution of fuel density. Our approach incorporates differentiable physics-informed terms in the loss function, including a mass-conserving fuel transport term a
The continuous advancements in AI and specifically deep learning architectures like ConvLSTM, AFNONet, and ViViT are enabling more sophisticated physics-informed models for complex systems.
Accurate and stable prediction of fuel density is critical for various industries, including energy, aerospace, and defense, directly impacting efficiency, safety, and operational costs.
The explicit integration of physics constraints into AI models leads to more robust and explainable predictions compared to purely data-driven approaches, potentially accelerating real-world adoption.
- · Energy sector
- · Aerospace engineering
- · Defense industry
- · AI/ML researchers
- · Traditional CFD modeling
- · Purely empirical modeling approaches
Improved predictive capabilities for fuel dynamics, leading to optimization opportunities in design and operations.
Reduced need for expensive physical experiments and simulations as PGML models become more reliable substitutes.
Accelerated development of novel fuels and propulsion systems due to faster and more accurate performance estimations.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG