The allocation of tugboat groups is critical for ensuring the timely and safe entry and departure of vessels in ports, while vessel scheduling determines the demand and timing for tugboat group deployment. This study introduces the collaborative optimization problem of vessel scheduling and tugboat group allocation (VT-COP) in seaports. An end-to-end deep reinforcement learning framework, integrating a graph neural network, is proposed to autonomously learn and optimize policies for solving the VT-COP. In this framework, the reinforcement learning agent simultaneously controls multiple actions, specifically assigning each vessel operation to a suitable tugboat group from a set of available options at each timestep. The multi-action nature of this problem is modeled as a multiple Markov decision process (MMDP). To address this, we introduce a novel multi-pointer graph network (MPGN) architecture and a bi-Proximal Policy Optimization (bi-PPO) algorithm. The MPGN architecture consists of two encoder-decoder modules that independently define action policies for vessel operations and tugboat groups, predicting probability distributions for different vessel operations and tugboat groups, respectively. The VT-COP is modeled as a disjunctive graph and utilizes a graph neural network to dynamically embed local states during the scheduling process. Computational experiments show that our approach significantly outperforms heuristic algorithms in terms of solution quality, with learned policies exhibiting strong generalization across various instances. Furthermore, tests on real-world scenarios confirm that the collaborative optimization of vessel scheduling and tugboat group allocation meets the operational efficiency requirements of modern seaports. By enabling more accurate and dynamic scheduling decisions, this study could reduce delays, optimize resource utilization, and enhance the overall throughput of port activities, offering a significant advantage in modern, high-demand port environments.

Due to the complex and dense traffic in ports and their surrounding sea areas, along with the diverse and varying sizes of ships, ship detection faces significant challenges. To address these challenges, a YOLOv8n-based ship detection method is proposed in this work. Firstly, based on YOLOv8n, two attention mechanism-CBAM and EMA-are integrated to improve attention allocation to ship target features in visible-spectrum imagery, thereby improving the feature extraction capability for multiscale ships. Secondly, considering the characteristics of overlapping ships and significant scale variations, a novel Loss function MPDIoU is adopted to address the inaccurate detection in scenarios with overlapping ships. Finally, a slim-neck lightweight neck structure is designed to reduce computational complexity while maintaining performance, thereby enhancing the inference speed of the network. Following these improvements, a ship target detection model named MSM-YOLOv8 was developed. Performance evaluation using the Seaships dataset demonstrates that MSM-YOLOv8 outperformed the baseline YOLOv8n in ship detection task, achieving an increase of 1.0 % in precision and 3.4 % in mAP@50–95, respectively. The proposed MSM-YOLOv8 was further validated both on Seaships dataset and ship images captured in real-world conditions, with the results confirming its effectiveness in accurately detecting and classifying various types of ship targets. In addition, experiments on the more complex ABOships dataset further demonstrate the robustness and generalization ability of the model. Therefore, the lightweight ship detection model proposed in this paper exhibits both theoretical significance and practical value in complex scenarios, and partially mitigates issues related to delayed detection and inaccurate classification of ship targets near ports.

Limitations are identified in the expressive capabilities of the deep feature extraction network employed in deep reinforcement learning (DRL), particularly in complex scenarios. Additionally, learning performance is negatively impacted by compound errors, with consideration given to the potential critical role of sample quality in learning outcomes. Therefore, how to improve the collision avoidance decision-making adaptive and effectiveness for autonomous ships navigating in various scenarios is crucial. To address these issues and overcome data acquisition difficulties, a method called rule-guided vision supervised learning (RGVSL) is proposed in this paper. Through a static collision avoidance decision-making task, a comparison is drawn between a deep feature extraction network and the Nature CNN in DQN, revealing shortcomings in the feature extraction of DRL. Additionally, a collision avoidance decision-making environment with the capability of generalization for maritime encounters is proposed, and the advantages of the method in terms of adaptability and learning cost are validated. Finally, it is demonstrated that the RGVSL method is equivalent to a DQN with γ set to 0, indicating a significant performance improvement without compound errors. Achieving an adaptive decision accuracy of over 90% in various encounter scenarios without retraining, this research substantially reduces learning costs. It can provide innovative and practical solutions for the technological development in the field of autonomous ship collision avoidance decision-making.

To address the issue of intelligent ship route planning, a ship planning method based on the improved D* Lite algorithm is proposed. Firstly, a navigation environment grid map is constructed using the acquired meteorological and hydrological datasets. The grids are divided into navigable and non-navigable according to navigation requirements, and a route planning model is built. Secondly, the heuristic function and the path function of the D* Lite algorithm are improved. The heuristic function is optimized and weighted, and a risk factor is introduced into the path function to enhance efficiency of path planning while maintaining a safe distance between the planned route and obstacles. Finally, by dynamically adjusting the search step length and the selectable directions of the D* Lite algorithm, the number of waypoints is reduced, and the voyage of the planned route is shortened, resulting in a smooth and collision-free route of ships. The effectiveness of the proposed algorithm is verified through three sets of simulation experiments. The simulation results show that the proposed method in this paper is more suitable for ship route planning and ship maneuvering in practice and can effectively avoid non-navigable grids while optimizing path length, path smoothness, and computation time, making the routes more aligned with actual navigation tasks.

Maritime Autonomous Surface Ships are increasingly becoming a topic of discussion due to the continuous improvement of ship’s intelligence levels. To meet the safety and economic requirements in a ship’s route design, a multi-objective intelligent hybrid algorithm based on the genetic algorithm and the greedy algorithm was proposed in this work. The marine environment model was gridded to a dynamic and static no-voyage zone. The initial continuous and random routes were generated by the greedy algorithm. The objective function, designed according to the multi-objective and multi-constraint, was solved by the genetic algorithm to generate the optimal route and corresponding speed distribution. The experimental results show that this method can ensure the safety of the ship and flexibly meet the needs of different decision-makers.

This paper realizes the simultaneous optimization of a vessel’s course and speed for a whole voyage within the estimated time of arrival (ETA), which can ensure the voyage is safe and energy-saving through proper planning of the route and speed. Firstly, a dynamic sea area model with meteorological and oceanographic data sets is established to delineate the navigable and prohibited areas; secondly, some data are extracted from the records of previous voyages, to train two artificial neural network models to predict fuel consumption rate and revolutions per minute (RPM), which are the keys to route optimization. After that, speed configuration is introduced to the optimization process, and a simultaneous optimization model for the ship’s course and speed is proposed. Then, based on a customized version of the A* algorithm, the optimization is solved in simulation. Two simulations of a ship crossing the North Pacific show that the proposed methods can make navigation decisions in advance that ensure the voyage’s safety, and compared with a naive route, the optimized navigation program can reduce fuel consumption while retaining an approximately constant time to destination and adapting to variations in oceanic conditions.

桥梁布设在通航水域的桥墩以及通航孔在一定程度上限制了船舶的航行,划定桥区水域对指导船舶安全航行具有重要意义.通过梳理当前桥区水域法规中船舶航行限制,分析划定桥区水域的影响因素;以桥梁涉水桥墩对航道内在航船舶的影响为依据,基于船舶领域理论设计了桥区水域宽度和纵向长度计算方法,可依据桥梁特点及航道水域特征对桥区水域进行划定,实现桥区水域划定的一桥一策;利用该方法对长江某大桥的桥区水域进行划定.结果表明,与传统桥区水域划定方法相比,该划定方法能够在保障船舶桥区航行安全的基础上,减少对航道内正常航行船舶的约束,同时可避免岸线资源的浪费.

立足"双一流"建设目标和新时代交通人才培养需求,以大连海事大学航海学院为例,分析水路交通运输专业研究生创新能力培养的现存问题,分别从培养目标、招生机制、课程学习、导师建设、科研实践和考评机制六个方面,提出水路交通运输专业研究生创新能力培养对策.

为研究避碰规则、无人水面艇(unmanned surface vessel,USV)运动学特点和海上交通复杂度等因素约束下的USV自主避碰技术,在分析初始动态窗口法的基础上,考虑《国际海上避碰规则》(International Regulations for Preventing Collisions at Sea,COLREGs)关于避碰行动时机、避让幅度、复航时机等方面的要求,建立融合避碰规则的动态窗口模型,设计融合避碰规则的动态窗口法.通过对比仿真实验验证该方法的可行性和有效性,具有一定的现实意义.

在大连海事大学大型船舶操纵模拟器的基础上,根据业内桥区水域的一般划定标准,设计出基于船舶操纵模拟器的桥区水域的划定方法,包括模拟区域电子海图的构建,模拟方案的确定和实施、模拟结果的统计和分析并得出最后的结论,并基于实例提出相应的桥区水域设计的方案和建议.

In order to solve the problem of ship route planning in the autonomous navigation and decision-making of ships, based on the electronic chart display and information system (ECDIS), the meteorological conditions of ocean-going ships were analyzed and the grid method was used to establish an environmental model. The voyage time, fuel consumption and navigation safety were taken as optimization goals, and a multi-objective ship route optimization model was established, then the non-dominated sorting genetic algorithm with elite strategy (NSGA-II) was applied in the optimization route searching to realize the solution of the optimal ship route.

Infrared cameras are more useful than visible light cameras in dark and foggy conditions; therefore, infrared imaging is becoming an increasingly popular subject of research. Feature extraction is an important aspect of image processing, but traditional convolutional neural networks (CNNs) trained on visible images cannot be used with infrared images. This study presents a method for retraining the Visual Geometry Group 19-layer CNN (VGG-19) to extract features from infrared images. First, a thermal image dataset was obtained from public datasets; this was then augmented by flipping, zooming, shifting, and rotating the images. Next, the architecture of the VGG-19 CNN was redesigned, and transfer learning was used to fine-tune the trainable layers. It was shown that the transfer-learned neural network could extract more information from infrared images than the original network could. To verify the validity of this method, it was also applied to the MobileNet, and the transfer-learned MobileNet also produced better results.

Aimed to the global demand for energy-saving and emission-reduction, and urgent need of shipping industry to cut down fuel costs, the wind-assisted ships are taken as an effective way for energy-saving and emission reduction of transoceanic crossing ships, this work focused on route optimization model to provide theoretical basis and technical support for wind-assisted ships. Firstly, this work analyses the route optimization of the wind-assisted ship, and two optimization objectives are defined. Secondly, the route optimization models for the minimum fuel consumption under the limited voyage time and the minimum voyage time on the condition of fixed main engine power were built separately, and the optimization algorithm based on simulated annealing was designed. At last, a 76,000DWT wind-assisted ship was taken as the experimental ship, and the feasibility and rationality of the model and algorithm was verified. As the simulation shown, the optimal route solved in this paper is effective in path planning problem of wind-assisted ship and has practical significance.

为了实现e-航海战略下海上搜救信息的数字化传输,提出e-航海环境下的海上搜救行动信息支持方法.首先,对海上搜救信息进行归纳总结,基于S-100通用海洋测绘数据模型创建了海上搜救信息要素类型,并利用统一建模语言(Unified Modeling Language,UML)对海上搜救信息进行了标准化建模,建立了海上搜救信息数据模型;然后,结合海上搜救的实际需求,搭建了包括岸端和船端两部分的信息支持平台,并采用可扩展标记语言(Extensive Markup Language,XML)对搜救信息进行编码使其能够在船、岸之间数字化传输;最后,对海上搜救信息数据模型和信息支持平台进行了仿真验证.仿真结果表明,该方法可使搜救信息在岸-船之间数字化传输,并在船端和岸端平台上以图形化的方式显示,能够为搜救行动提供决策支持.

Due to many limitations and deficiencies in the onshore weather routing, the onboard weather routing based on ant colony optimization (ACO) is proposed in this paper. Firstly, this work analyses the similarities and differences between weather route optimization and TSP, proposes a more appropriate heuristic function which makes ants tend to search for grids nearer to the destination. Secondly, an onboard weather routing algorithm based on ant-cycle model is established, including constraint criteria, path search strategy and smoothing strategy. At last, the history meteorological data was used to simulate the forecast information, and the feasibility and rationality of the model and algorithm is verified based on an experimental ship. As the simulation shown, the final optimal route solved in this paper can be in line with the actual situation of navigation and having practical significance.

The campus informatization is the trend of the future development of school education, and mobile information service is the vanguard of this tide, WeChat as the most popular mobile information service products, has attracted great attention. At the same time, the internet teaching method has become the trend of reform in vocational colleges. From the practical, to the Career Academy Tianjin maritime college as an example, presents the design idea of the system of open information query platform based on WeChat public query system design, system design, training query from teaching makes corresponding discussions.

With the continuous deepening of the vocational education teaching reform, internet teaching technology plays an increasingly important role in the practice of vocational education. This paper takes the course of "navigation" navigation technology specialty as an example, developed the college students often use mobile phone software in the social media platform, designed the function of auxiliary teaching media platform, custom menu, background settings, graphic push and other projects, and summarizes and analyzes the experience of the construction of media platform.

根据风帆空气动力学特性、航线、季风风场,提出将典型风帆(翼型帆)的能量转换效果指标——转换系数等值线叠加在风能密度图上的方法,对典型航线上的可用风力资源进行分析.首先,推导风能密度和翼帆推进力公式,提出推进功转换系数这一参数来评价翼帆对不同特征的风能的利用能力;然后,结合风能特征分布图和转换系数等值线,提出可用风力资源分析方法;最后,以宁波一中东航线为例,使用ECMWF的2014年风场数据对其可用风力资源进行分析计算.结果显示:船舶航向不同时,中国南海冬季、夏季的可用风力资源差别较大,夏季应采用W-E航向,冬季应采用E-W航向.孟加拉湾、阿拉伯海区域可用风力资源特征相似,夏季采用W-E航向能够更高效地利用风能.

As a result of a global call for energy-saving and emission-reduction strategies as well as an urgent need to reduce the shipping cost of transoceanic crossings, this paper proposes a route that minimizes the time for such crossings and provides technical support to efficiently utilize wind power based on existing research for wind-assisted ships. To begin, the ocean winds around the ship route were analyzed, and the different influences on traditional ships and wind-assisted ships were listed for various wind speeds and directions. The number of waypoints of a route was subsequently calculated, and a model of the optimal ship route was then built based on the fixed power output of the main marine engine. A solution algorithm based on simulated annealing was then presented to determine the optimal wind-assisted ship routes by minimizing the travel time. Finally, a 76,000-DWT wind-assisted cargo ship was designated as the experimental ship, and the optimization model and its algorithm were simulated to generate an optimized wind-assisted route. The simulation indicated that the speed of a ship equipped with wind propulsion increases, which significantly reduces the travel time and fuel costs over the optimized route, despite the increased distance of this route. Thus, the route optimization algorithm designed in this study can be applied to optimize the routes for wind-assisted ships and theoretically guide further studies of wind-assisted projects.