To address the challenges in dynamic modeling of Autonomous Underwater Vehicle (AUV), this paper proposes a black-box identification method for nonlinear systems based on deep convolutional neural networks, taking into account the nonlinear characteristics of the AUV's six-degree-of-freedom (6-DOF) motion. First, the frequency corresponding to the maximum amplitude of the rudder signal is extracted and used as a threshold for Variational Mode Decomposition (VMD) denoising. This reduces noise in the experimental data of the AUV model and resolves the issue of difficult parameter tuning in VMD decomposition. Then, a black-box model for the nonlinear system is constructed using Bidirectional Long Short-Term Memory (BiLSTM) and Attention mechanisms, with the Adam optimization algorithm employed to solve the AUV 6-DOF motion model. Finally, the AUV model data are used for model training and predictive validation, and the results are compared with modeling methods such as CNN-LSTM, CNN-BiLSTM, and CNN-LSTM-Attention to analyze the velocity, Euler angles, and trajectory of AUV motion. Experimental results show that, compared to the CNN-LSTM model, the proposed method improves the Root Mean Square Error (RMSE), the coefficient of determination (R²), and the Symmetric Mean Absolute Percentage Error (SMAPE) by 79.29%, 3.84%, and 74.41%, respectively, validating the feasibility and effectiveness of the proposed dynamic modeling approach. This method provides an alternative strategy for precise obstacle avoidance and autonomous navigation of underwater vehicles.

As an important component of the waterway transportation system, Ro-Ro passenger ship transportation plays a significant role in inland river, coastal, and even cross-strait transport services. In recent years, collisions involving Ro-Ro passenger ships have occurred from time to time. To mitigate the losses caused by such accidents, this paper proposes an emergency decision-making model for Ro-Ro passenger ship collisions based on a fuzzy Bayesian network. The identified emergency decision variables for RoPax ship collisions are fuzzified by introducing fuzzy logic. Combined with improved IF-THEN rules, confidence rule bases are established and then converted into a conditional probability table, thereby constructing a complete Bayesian network inference structure. Ultimately, the optimal emergency decision scheme is determined through utility value evaluation. The results demonstrate that the proposed emergency decision-making model is effective and feasible, aligning with practical application requirements. This study provides ship decision-makers with a reference basis for emergency response in the event of a RoPax ship collision.

To support the autonomous navigation of cargo-carrying vessels with specific time and position requirements, research on high-precision trajectory tracking control is necessary. In response to the limited existing studies on cargo vessels and the insufficient consideration of actuator characteristics-where thrust and torque are often treated as directly controllable inputs, leading to limited practical feasibility-a control method combining a virtual vessel leader and an Integral Line-of-Sight (ILOS) approach is proposed. This method uses a propeller speed prediction algorithm to synchronize the real vessel with the virtual vessel and employs speed feedback correction to compensate for disturbances. To improve tracking accuracy, the relative positions are utilized to determine the desired heading through the improved ILOS method, thereby reducing the problem to one of course keeping. Ultimately, vessel trajectory tracking is achieved. Simulation results show that the controlled vessel accomplishes trajectory tracking under disturbance, with a steady-state error of less than ±0.5 m, an error convergence time of less than 50 s, an 86% reduction in rudder jitter, and 71% reduction in propeller speed jitter. The proposed control method is straight forward, demonstrating high performance, can serve as a valuable reference for engineering applications.

Maritime fire incidents pose a significant threat to the safety of ships, with human factors being the primary cause of these accidents. Accurately identifying the emotional changes of crew members in maritime fire scenarios is of great significance for enhancing their firefighting capabilities. Virtual reality technology is employed to simulate maritime fire scenes and collect Electroencephalogram (EEG) signals from multiple subjects. The EEG signals are preprocessed and decomposed into sub-signals of different frequency bands using discrete wavelet transform. Three features, including mean absolute value, standard deviation, and root mean square, are extracted from each sub-frequency band to establish a feature set. Multiple machine learning models suitable for emotion recognition are constructed, and the models are evaluated using metrics such as precision, accuracy, and F₁ score. Experimental results show that the support vector machine classification model performs the best, with an accuracy of 87.97%, which significantly improves the three-class classification problem of crew members' fear emotions in maritime environments. Combining virtual reality technology with EEG emotion recognition techniques can effectively induce and identify crew members' fear emotions in fire scenarios. This method is beneficial for assessing and improving the emergency response capabilities of crew members in firefighting training.

Currently, ship navigators can assess flow patterns using basic instruments and adjust maneuvering strategies accordingly. Access to detailed flow field data of a waterway can provide valuable information and early warnings for ships transiting the area. This study analyzes surface flow in the waters near Jianghan Bridge, captured by video. By employing Large-Scale Particle Image Velocimetry (LSPIV), a method is developed to measure surface flow velocity in the navigation channel, enabling analysis of surface flow characteristics and acquisition of surface flow field data. The obtained flow field data are validated through comparison with optical flow methods and Acoustic Doppler Velocimetry. Results demonstrate that the proposed surface flow velocity measurement method can effectively capture detailed flow pattern characteristics of surface currents in the study area. This approach provides data support for navigation and path planning of both conventional ships and smart ships utilizing big data, contributing practical value to the enhancement of maritime safety and operational efficiency.

With the rapid development of global shipping, port cargo volumes are increasing significantly, leading to growing issues of vessel congestion and delays, which in turn severely constrain port operations. In response to the challenges posed by surging cargo volumes, ship congestion, and aggravated pollution, this paper proposes a Time-Berth & Pollution (TB&P) model. The model takes the difference between the actual and expected time in port as the objective function, aiming to minimize operating costs and pollution emissions, subject to constraints related to time, space, and machinery/equipment. To solve the TB&P model, an improved version of the basic Beluga Whale Optimization (BWO) algorithm is developed, termed the Opposition Learning Beluga Whale Optimization (OBWO) algorithm. The feasibility and superiority of the proposed model and improved algorithm are verified through case data from a port. Results demonstrate that, compared with traditional models, the established model significantly reduces the extent of ship delays and mitigates water pollution in the port area. Furthermore, the proposed OBWO algorithm exhibits enhanced stability and accuracy relative to other selected algorithms.

The arrival and handling times of vessels are subject to significant uncertainty. Triangular fuzzy numbers, characterized by upper and lower bounds and a most likely value, provide an effective means of representing such imprecise information. In this context, this paper first establishes a fuzzy integer programming model for berth allocation at container terminals, aiming to minimize the total departure delay time of vessels. An improved Multi-Verse Optimizer (MVO) algorithm is then proposed to solve the model, incorporating solution repair and breakout strategies. Comparative analysis shows that, in contrast to deterministic berth allocation schemes, the proposed fuzzy berth allocation approach demonstrates notable advantages in reducing total departure delay time and exhibits greater effectiveness in handling uncertainty. Moreover, the improved MVO algorithm achieves solution speed improvements of 59.9%, 44%, and 26.1% in small, medium, and large scale experiments, respectively, compared to the standard multi-verse optimizer. These results indicate that the proposed algorithm can efficiently solve the fuzzy integer programming model for berth allocation and offers valuable decision-making support for addressing berth allocation problems under fuzzy uncertainty.

In the context of carbon emission reduction, this study constructs a bi-level planning model for port microgrid investment and deployment, with the government as the upper-level decision-maker and the port area as the lower-level follower. The model incorporates the interests of the port area, including berthed vessels, and aims to maximize environmental benefits while minimizing the total cost of the port area over the planning period. Using the Column and Constraint Generation (CCG) algorithm, the optimal investment and operation strategy for the port area during the planning horizon is derived. The study analyzes the deployment of the port microgrid system under varying incentive budgets and evaluates the resulting environmental benefits, comparing the effectiveness of different incentive strategies. The results demonstrate that a hybrid incentive strategy can significantly enhance investment motivation in port microgrid systems, thereby effectively fostering innovation in the energy structure of the port region and accelerating the emission reduction process.

To foster a new development paradigm, the Yangtze River Economic Belt serves as a crucial hub connecting the domestic and international circulations. As the leading port and core node of the Yangtze River Economic Belt, Shanghai Port's strategic role as one of the most critical infrastructure has become increasingly prominent. The Shanghai International Shipping Center is facing a bottleneck of spatial resource constraints in upgrading its service capacity. This paper evaluates port operational efficiency by constructing models for port service intensity, waiting probability, and queuing theory. It quantifies the impact of resource constraints using a ship loss rate model under different system capacities, while comparing development strategies of typical domestic and foreign ports to analyze the core challenges confronting the Shanghai International Shipping Center. Model analysis indicates that the structural contradiction between port service demand and spatial resource supply has become a key constraint on its high-quality development. Looking ahead, the expansion and upgrading of port-shipping resources and optimizing spatial resource allocation will enable the Shanghai International Shipping Center to effectively unleash the advantages of direct river-sea transportation, significantly improve port service capacity and the navigation efficiency of the Yangtze River Golden Waterway, and reduce the comprehensive logistics costs of the whole society. This development path will not only promote the green transformation and digital-intelligent upgrading of the shipping industry but also strengthen the country's supply chain security and expand high-level opening-up, laying a solid foundation for the long-term and sustainable development of the Shanghai International Shipping Center.

Accurate forecasting of port container throughput is of great significance for port operators and government administrations in making scientific decisions. Existing forecasting methods, however, often pay insufficient attention to short-calendar-time PCT and exhibit limited accuracy in handling nonlinear and non-stationary fluctuation series. This paper takes the container throughput of Shanghai Port as the research object and proposes a novel deep learning model based on secondary decomposition using CCVMD and STL. Using the correlation coefficient as a reference, variational mode decomposition is first applied to the original time series. Subsequently, a secondary decomposition divides the data into seasonal, trend, and residual components. An algorithm-optimized long short-term memory neural network is then employed to predict each component separately, and the final prediction results are aggregated. Experimental results show that the combined decomposition model with data preprocessing significantly outperforms other models in PCT forecasting. The proposed model achieves a mean absolute percentage error of 0.021 703, a root mean square error percentage of 0.026 852, and a mean absolute error percentage of 0.022 14, indicating superior overall performance compared to 12 benchmark models and several models from prior studies. Furthermore, the secondary decomposition approach demonstrates enhanced reliability in tracking extreme values, removing and reducing noise, and improving interpretability.