A deep learning-based method for predicting ship dynamic stability in real-sea conditions

To address the limitations of the IMO second-generation intact stability assessment method—especially its inability to capture the spatiotemporal variability of real ocean environments and its limited accuracy in evaluating ship instability risk—this study proposes a sensitivity index prediction approach based on real-sea environmental conditions, aiming to enhance the navigational safety of container ships.

By integrating the Second Generation Intact Stability Criteria with deep learning techniques, an accurate regional sensitivity index prediction model was developed. Typical container vessels with recorded stability accidents were used as case studies to systematically validate the predictive accuracy of the sensitivity indices.

The proposed model significantly improves temporal and spatial resolution compared to the IMO method, enabling accurate identification of the location and timing of instability risks. The predicted results show high consistency with actual risk distributions without overestimating risks. Compared to the IMO method, the model better captures the variability of stability risks under complex sea conditions.

The developed sensitivity index prediction model fully accounts for the spatiotemporal variability of real ocean environments and enables rapid prediction of second-generation ship stability under realistic sea conditions. It improves both the accuracy and practical applicability of risk forecasting, offering valuable decision-support information for container ship safety assurance and route planning. This research holds significant implications for the protection of ships and cargo at sea.