Physics-informed neural network modelling for deep-water linear wave propagation

Key Points

• Effective modeling of deep-water wave propagation is achieved using physics-informed neural networks, allowing for greater accuracy.
• The study shows that this approach can reduce computational costs significantly in numerical simulations involving complex fluid dynamics.
• The modeling technique incorporates physical laws directly into the neural network structure, providing insights into linear wave behavior.
• Implications include improved understanding of ocean dynamics and expanded capabilities for forecasting wave patterns in real-time.