Add BrightSurf on Google Email

Hybrid deep learning framework enables real-time, high-precision prediction of midship bending moments in severe seas

Researchers developed a three-stage autoregressive neural network architecture to predict midship bending moments in real-time, offering continuous probabilistic confidence bounds and probabilistic safety intervals. The model achieved high accuracy, with an average R² of 0.9533 and a 94.22% coverage rate at the 95% confidence level.

SourceTsinghua University Press·JournalOcean·DateAug 20, 2026

Pusan National University researchers build a numerical algorithm to study continuous ice-breaking

Researchers at Pusan National University have created a new algorithm that can accurately predict ice resistance and fracture points for ships navigating through the Arctic shipping routes. The model uses an elastic material approach, allowing it to study continuous ice-breaking processes, which is essential for efficient navigation.

SourcePusan National University·JournalOcean Engineering·TypeComputational simulation/modeling·DateMar 30, 2023