Foggy weather significantly degrades ship visibility and image quality, posing serious risks to navigation safety. Enhancing the dehazing performance of ship navigation images is therefore of great importance. To address the insufficient fog removal and poor detail restoration of existing dehazing methods in maritime scenarios, this study proposes an end-to-end ship image dehazing method that integrates an improved CycleGAN with attention mechanisms. A Squeeze-and-Excitation （SE）channel-attention module is introduced to aggregate feature maps, compress spatial information, and strengthen the network's ability to learn global representations. Multi-scale channel fusion is achieved through skip connections, which not only reduces computational complexity but also enables the model to better capture fog characteristics under complex atmospheric conditions and to process ship targets of different sizes. Furthermore, a Channel Attention module is incorporated to enhance feature selection and improve the restoration of ship contours and fine structural details. Quantitative evaluations and real fog-navigation experiments confirm the robustness of the proposed method, demonstrating consistent improvements over existing dehazing approaches across all tested metrics and navigation scenarios.

This paper presents a contrast-enhancement-based dehazing algorithm to address detail loss, dim brightness, and color distortion in dehazed images of offshore tugboat sailing scenes with large sea-sky regions. First, atmospheric light estimation is optimized using quadtree segmentation to locate the light source in regions with minimal local pixel variance. Then, mean squared error contrast preserves image details, and a contrast overall cost function combined with an information loss function is used to find optimal transmission, enhancing contrast and making the sky region clearer. A fast guided filter further refines the transmission map, reducing block artifacts and maintaining real-time performance while restoring image authenticity. Finally, adaptive histogram equalization preserves contrast information in the sky, avoiding over-bright or over-dark areas. Experiments show that the image obtained using the proposed algorithm improves structural similarity, peak signal-to-noise ratio, and mean squared error by 15. 94%, 11. 46%, and 25. 82%, respectively, compared with the OCE method, while preventing color cast and halo effects, enhancing sea-sky boundary clarity, and meeting real-time requirements for restoring a realistic maritime environment.

Ship fuel consumption prediction plays a crucial role in navigation decision-making and the intelligent evaluation of energy efficiency, particularly for future Marine Autonomous Surface Ships (MASS). This study leverages an onboard measurement and data acquisition system installed on a 28, 000 DWT bulk carrier operating on global routes. With the system, navigation-related data from 2010 to 2016 across different sea areas, loading conditions, and meteorological and sea states were collected and analyzed, including ship speed, course, sway, main engine speed, and environmental parameters. Using real-time inputs such as wave height, wave direction, speed, wind speed, pitch angle, main engine power, and main engine speed, a fuel consumption prediction model was developed based on the lightGBM algorithm. The performance of this model was compared with other machine learning approaches, including Support Vector Regression（SVR）, Long Short-term Memory （LSTM）, Gated Recurrent Unit （GRU）, Artificial Neural Network （ANN）, and Extreme Gradient Boosting （XGBootst）. Results show that the proposed model achieves superior performance, with RMSE reduced by 7.26%, MAE reduced by at least 2.62%, R² increased by 0.23%, and runtime shortened by 73.76%. Furthermore, the navigation data was divided into four subsets based on actual loading conditions to further validate the generalization capability of the lightGBM model. The results indicate that the proposed LightGBM model provides an effective solution for predicting ship fuel consumption, striking a balance between accuracy and computation efficiency. This study also provides a valuable reference for selecting optimal fuel consumption prediction methods in comparable vessel types.

This study develops a coverage path planning method for multi-UAV maritime search and rescue (MSAR) missions under dynamic ocean conditions and time-critical constraints, aiming to balance search efficiency and resource allocation. Firstly, a grid-based regional decomposition approach is adopted to discretize complex maritime environments into visual planning cells, while a Gaussian Mixture Model (GMM) is employed to construct a prior target-drift distribution and generate a probabilistic map for path guidance. Secondly, for multi-UAV coverage planning, an improved Multi-Objective Particle Swarm Optimization (MOPSO) algorithm is developed to jointly optimize task allocation, path safety, coverage of high-priority areas, and energy consumption control. Thirdly, to enhance global search capability and convergence performance, the algorithm incorporates a Sigmoid-based adaptive inertia weight strategy, a two-level elite-guided crossover strategy, and a constraint-penalty mechanism. Finally, three UAVs were deployed to conduct simulation tests over MSAR regions of various shapes. Results show that, compared with classical baseline algorithms, the proposed method achieves up to 30. 27% improvement in cumulative detection probability, 82. 5% improvement in workload balance, and 1. 28% reduction in total path length within the first 50 steps, demonstrating its effectiveness and practicality for improving MSAR efficiency and coordination.

Hydrodynamic wake is one of the important physical fields for the non-acoustic detection of submarines, and its accurate prediction is difficult due to the complexity of the marine environment. The Volume of Fluid (VOF) method is used to numerically simulate the stratified flow field of a submarine under direct navigation, and the influence laws of the speed, dive depth, and seawater stratification on the free-surface wake features and internal-wave wake features are obtained. Canny edge detection operator theory is applied to extract the wake profile features, and the wake angle and transverse wave tensor angle are measured. The results show that the submarine's wake does not present a typical "V" wave when the Fr number is low at a large dive depth, but gradually changes from a "spot" to a parallel wave with the increase of the Fr number, and eventually evolves into a distinct "V" wave. The strong stratification of seawater enhances free-surface perturbation, making the free-surface wake characteristics more pronounced. These results provide a basis for image recognition of hydrodynamic wake characteristics of submarines.

To address the limitations of the conventional square grid in ship path planning—such as insufficient safety margins, low search efficiency, and poor adaptability to ship maneuvering characteristics—this study proposes a ship path planning method based on a regular hexagonal grid and an improved A^* algorithm. A hexagonal grid neighborhood model is constructed according to the geometric properties of the regular hexagonal grid and the required ship-obstacle safety clearance, together with an encoding scheme suitable for hexagonal cells. A ship motion cost model is developed by incorporating ship inertia and turning constraints. Based on this model, the traditional A^* algorithm is improved by optimizing the heuristic function and introducing a turning-penalty mechanism, thereby forming a hexagonal-grid-based search algorithm for ship path planning. Comparative experiments show that, compared with the square-grid 8-neighborhood method, the proposed method shortens the path by 5.51% and reduces the number of search nodes by 30. 7% ; compared with the 4-neighborhood method, the path length is reduced by 17. 0% and the number of turning points decreases by 38. 6%. The paths generated on the hexagonal grid are smoother and more consistent with ship maneuvering characteristics. Key words:intelligent navigation; ship path planning; hexagonal grid; improved A^* algorithm; ship maneuverability

In the automated container terminal yard with a vertical layout, the collaborative scheduling and resource allocation of non-crossing Automated Stacking Cranes (ASC) are crucial for improving operation efficiency. The traditional fixed buffer block strategy is prone to causing path conflicts and resource competition bottlenecks due to the uneven distribution of tasks, resulting in limited operation efficiency. To address this issue, this paper proposes a joint optimization framework for the scheduling of two ASCs and the decision-making of flexible buffer blocks. The aim is to achieve efficient collaboration of yard resources through a dynamic task-block matching mechanism and a safety time interval constraint model. Firstly, a mixed-integer programming model is constructed. With the goal of minimizing the task completion time, it couples the path planning of ASCs, task sequences, and the dynamic allocation of buffer blocks. Secondly, a multi-chromosome encoding genetic algorithm is designed, and the sorting crossover and Gaussian mutation strategies are adopted to enhance the solution efficiency for large-scale instances. Numerical experiments show that compared with the single fixed buffer block strategy, the flexible buffer block mechanism can reduce the average task completion time by 13.9%, verifying the adaptability and accuracy of the dynamic allocation. This study provides theoretical support and decision-making basis for the resource allocation in automated terminal yards.

The Shipborne Helicopter Landing Platform (SHLP) provides an effective solution for ensuring the safe landing of shipborne helicopters by compensating for ship motions induced by wind, waves, and currents. To address the stabilization control problem of the SHLP subject to compound disturbances caused by ship motions, dynamic uncertainties, and load perturbations, a preset-time robust adaptive stabilization control method is proposed. First, a novel barrier Lyapunov function is designed, which allows the convergence time for regulating the supporting surface of the SHLP to the desired horizontal position to be preset in advance. Then, an adaptive compound disturbance estimator is constructed to achieve online estimation of the compound disturbances acting on the SHLP. Finally, by integrating the backstepping design approach with a projection algorithm, a preset-time robust adaptive stabilization control law for the SHLP is developed. Numerical simulation results demonstrate the effectiveness and robustness of the proposed control method under compound disturbance conditions.

To address the difficulty in quantitatively measuring and recognizing the similarity of multi-ship encounter scenarios, a similarity-based recognition method for multi-ship encounter scenarios based on topological graph sequences is proposed. First, multi-ship encounter scenarios are extracted from Automatic Identification System (AIS) data and represented using a topological graph sequence-based model that characterizes ship interaction relationships. Then, a two-stage similarity recognition algorithm is designed to calculate the similarity between graph sequences and identify similar encounter scenarios. Taking the waters of Ningbo-Zhoushan Port as a case study, 2, 898 multi-ship encounter scenarios are extracted from one month of AIS data, and two typical scenario types with higher proportions are selected for experimental validation. The recognition results are comparatively analyzed based on encounter feature parameters. Experimental results show that the identified similar scenarios exhibit dynamic evolution characteristics highly consistent with those of the original scenarios, and the proposed method can effectively recognize multi-ship encounter scenarios with similar encounter relationships. This demonstrates the feasibility and effectiveness of the proposed method for similarity measurement of multi-ship encounter scenarios. The findings can provide a reference for collision avoidance decision-making and encounter risk analysis in multi-ship encounter scenarios.

The application of green ship technologies, such as hull form optimization, has been widely studied for improving energy efficiency during ship operation. However, environmental evaluations of the shipbuilding and scrapping stages remain insufficient, making it difficult to assess the lifecycle environmental impacts of these technologies holistically. To address this gap, this study selects a 150, 000-ton shuttle tanker as the research object. Experimental measurements and statistical analyses were conducted to quantify differences in energy and material consumption over the ship's lifecycle resulting from hull form optimization. Using the Life Cycle Assessment （LCA）method implemented in SimaPro V9. 6 software, an environmental impact assessment was performed to analyze the effects of hull form optimization on Global Warming Potential （GWP）and its contribution to carbon emission reduction. The results indicate firstly that the carbon reduction contribution of hull form optimization is highest in the shipbuilding stage, followed by the operation stage, and lowest in the scrapping stage. This suggests that a comprehensive evaluation of green ship technologies should account for not only the operational phase but also the construction and dismantling phases. Furthermore, while both LCA and the Energy Efficiency Design Index （EEDI）methods show broadly consistent trends in assessing the carbon reduction effect of hull form optimization during operation, the LCA results are more conservative. This discrepancy arises partly from differences in operational condition assumptions and the fact that the carbon emission factors in the Ecoinvent-3 database, commonly used in LCA, do not fully account for fuel combustion processes.