Flow-induced vibration (FIV) is a phenomenon in which the flow passing through a structure exerts periodic forces on the structure. Most studies on FIVs focus on suppressing this phenomenon. However, the Marine Renewable Energy Laboratory (MRELab) at the University of Michigan, USA, has developed a technology called the vortex-induced vibration for aquatic clean energy (VIVACE) converters that reinforces FIV and converts the energy in tidal currents to electrical energy. This study introduces the experimental data of the VIVACE converter and the associated method using deep neural networks (DNNs) to predict the dynamic responses of the converter. The DNN was trained and verified with experimental data from the MRELab, and the findings show that the amplitudes and frequencies of a single cylinder in the FIV predicted by the DNN under various test conditions were in good agreement with the experimental data. Finally, based on both the predicted and experimental data, the optimal power envelope of the VIVACE converter was generated as a function of the flow speed. The predictions using DNNs are expected to be more accurate as they can be trained with more experimental data in the future and will help to substantially reduce the number of experiments on FIVs.

Plates and curved plates are essential components in ship construction. In the design stage, the methods used to evaluate the ultimate strength required to confirm the structural safety of plates include prediction through analytical methods, finite-element analysis (FEA), and empirical formulas. However, with nonlinear buckling, the results of the empirical formula and the FEA differ for small flank angles (1~9). As a result, the prediction of the nonlinear ultimate strength of flank angle (1~9) plates still requires significant computation time and cost. To compensate for this, this study performed an ultimate strength prediction method utilizing a deep neural network together with the 4050 curved plate analysis. In addition, this paper presents the analysis results of the nonlinear finite-element method and the geometric shape and ratio of curved plates as training data. Based on the results of this study, designers can more efficiently design appropriate curved plate members by considering the ultimate strength.

Accurate prediction of ship roll motion is essential for safe and autonomous navigation. This study presents a deep learning framework that estimates both roll motion and epistemic uncertainty using Monte Carlo (MC) Dropout. Two architectures, a Long Short-Term Memory (LSTM) network and a Transformer encoder, were trained on HydroD–Wasim simulations covering various sea states, speeds, and damage conditions, and validated with real voyage data from two ferries, showing complementary performance, where LSTM achieved higher accuracy and Transformer provided more reliable confidence intervals. Model performance was evaluated by mean squared error (MSE), prediction interval coverage probability (PICP), and prediction interval normalized average width (PINAW). The LSTM achieved lower MSE, showing superior deterministic accuracy, while the Transformer produced higher PICP and wider PINAW, indicating more reliable uncertainty estimation. Results confirm that MC Dropout effectively quantifies epistemic uncertainty, improving the reliability of deep learning–based ship motion forecasting for intelligent maritime operations.

The initial response to a marine accident can play a key role to minimize the accident. Therefore, various decision support systems have been developed using sensors, simulations, and active response equipment. In this study, we developed an algorithm to predict damage locations using ship motion data with bidirectional long short-term memory (BiLSTM), a type of recurrent neural network. To reflect the low frequency ship motion characteristics, 200 time-series data collected for 100 s were considered as input values. Heave, roll, and pitch were used as features for the prediction model. The F1-score of the BiLSTM model was 0.92; this was an improvement over the F1-score of 0.90 of a prior model. Furthermore, 53 of 75 locations of damage had an F1-score above 0.90. The model predicted the damage location with high accuracy, allowing for a quick initial response even if the ship did not have flood sensors. The model can be used as input data with high accuracy for a real-time progressive flooding simulator on board.

Considering the nature of marine accidents, even a single accident can result in significant damage to the environment and property, as well as loss of life. Therefore, the initial response should be rapid and accurate, and various decision support systems have been developed to achieve this. Research on simulating progressive flooding on board immediately after an accident is being actively conducted, but this requires high levels of computing power. In this study, a methodology for converting simulated ship motion data into a ship motion database is presented. The model of a training ship from the Korea Institute of Maritime and Fisheries Technology and KRISO in-house code SMTP was used for ship motion computations. The short-time Fourier transform was used to convert time-series motion data into a spectrogram motion database. A methodology for deriving a predicted location of the damage center is presented. The candidate locations of the damage centers were obtained by comparing the root mean square error values of the ship motion database from the simulation and real-time ship motion data. Finally, a probability function was suggested to confirm the predicted location of the damage center. Using 100 randomly selected test cases, our method showed 95% accuracy.