Against the backdrop of advancing national low-altitude economy strategies, this research examines the factors and mechanisms influencing user acceptance of two differentiated low-altitude air travel services: the fixed-route shuttle (public service type) and 'Feidi' (market-oriented service type). By refining the integrated Technology Acceptance Model–Theory of Planned Behavior (TAM–TPB) framework, a multidimensional "technology-institution-cognition" analytical framework was constructed and empirically validated using a dual-scenario structural equation modeling approach based on questionnaire survey data. The findings reveal that fixed-route shuttle services exhibit a technology-function-dominated decision-making logic, where technology trust not only directly drives usage intention but also indirectly enhances acceptance by improving perceived usefulness and reducing risk perception. In contrast, 'Feidi' services follow an experience-driven decision-making pathway, where technology trust indirectly influences usage intention through attitude, with user decisions relying more heavily on perceived behavioral control and subjective experience. Drawing from the underlying user decision mechanisms, differentiated strategic guidance is provided for manufacturers' technology development priorities, operators' user experience optimization, and regulatory authorities' institutional design, thereby offering theoretical foundation and operational direction for promoting the industrial application of low-altitude air travel services.

The IHO S-98 standard defines the interoperability between electronic navigational charts (ENCs) and multi-source hydrographic data as a future application trend. This study aims to explore intelligent techniques for generating dynamic depth contours by integrating ENC products with real-time water level data. In scenarios where S-102 bathymetric surface data are incomplete, the sparse depth data contained in ENC products are insufficient to directly support high-reliability dynamic depth computations. To address this limitation, we propose a deep learning-based super-resolution solution. Specifically, an improved TSE-EDSR model is employed to reconstruct a high-fidelity digital elevation model (DEM) of the seabed from sparse chart soundings. By further integrating real-time water level data, dynamic depth contours are generated. Experimental results demonstrate that the proposed method produces DEMs and dynamic depth contours with significantly higher accuracy and morphological authenticity compared to traditional interpolation approaches. This work provides technical support for S-98 interoperability applications and holds substantial theoretical and practical value for advancing the development and application of next-generation electronic navigational charts.

In response to the industry pain point of high cost of ocean freight empty container repositioning, an innovative combination of shipping company cooperation mechanism and free detention time strategy is proposed. Taking into account four channels for obtaining empty containers, namely container repositioning, storage, rental, and exchange, a nonlinear programming model is constructed with the goal of minimizing the total cost of empty container repositioning. By designing an approximate dynamic programming algorithm based on  greedy strategy, the multi-period decision-making and nonlinear coupling problems that traditional methods are difficult to handle have been solved, and the optimal decision for the length of free container period in different scenarios has been provided. Research has found that setting a reasonable free container detention period can reduce container rental costs and optimize container usage costs under cooperation with shipping companies. Sensitivity analysis shows that when the supply of empty containers exceeds the demand, the total cost of empty container management for shipping companies will reach its optimal level. For every 33% increase in inland container volume or 1 day extension in transportation time, the free detention period needs to be shortened by 1-2 days to achieve cost optimization. This collaborative decision-making framework helps improve the efficiency of empty container resource turnover in shipping companies and provides a basis for determining the length of free detention period for shipping companies. 

To adapt to changes in the shipping market and meet the demands of low-carbon transition, shipping companies need to make operational decisions regarding fleet deployment at regular intervals. To enhance the decision-making efficiency of shipping companies, a joint optimization model for fleet deployment and speed in liner shipping networks is formulated, considering multiple factors of vessel selection, cargo allocation across routes, and various green shipping measures. Considering the inherent sequential decision-making characteristics of this model, a two-stage interactive algorithm incorporating linear transformation and cascading increment strategies is proposed for efficient solution. The accuracy and computational efficiency of this algorithm are validated through algorithm comparison and analysis. Compared to the enumeration method, it obtains the same optimal solution and improves computational efficiency by approximately 99%. Finally, taking the Trans-Pacific shipping network as a case study, we investigate optimal fleet deployment strategies. The results demonstrate that utilizing vessels with larger container capacity while increasing sailing speeds significantly enhances operational profitability of carriers. The research results enable shipping companies to dynamically adapt to market fluctuations and formulate operational strategies with enhanced efficiency.

In modern industrial systems, mechanical vibration monitoring technology plays a crucial role in ensuring equipment safety and preventing failures. This study proposes a highly sensitive liquid metal-based vibration sensor based on a triboelectric nanogenerator (TENG) for real-time vibration monitoring of marine mechanical equipment. The sensor consists of conductive fabric, Fluorinated Ethylene Propylene (FEP) film, and liquid metal, utilizing the triboelectric effect to convert mechanical vibrations into electrical signals. Experimental verification shows that this sensor exhibits excellent linear voltage response (R 2=0.995) within the dynamic acceleration range of 5 to 50 m/s⊃2;, with a sensitivity of 0.218 V·m⁻⊃1;·s⊃2;. It also has good durability, and the signal attenuation can be ignored after 21,600 acceleration fatigue tests. It was ultimately successfully applied to the vibration monitoring of ship air compressors. Compared to traditional piezoelectric and electromagnetic sensors, this technology offers advantages such as high sensitivity, electromagnetic interference resistance, and flexibility to adapt to harsh environments, providing a novel solution for condition monitoring of equipment in demanding operational scenarios such as marine applications.

The safe and stable operation of marine engines is critical to national security and maritime traffic safety. While numerous deep learning methods have been extensively studied for intelligent fault detection, the complex operating conditions of marine engines—such as non-stationary states including variable loads—often lead to prevalent domain shift problems in practical fault diagnosis tasks. This significantly degrades the performance of conventional deep learning approaches. Using a specific marine engine as a case study, we constructed partial-set fault diagnosis scenarios under varying operating conditions. To address the challenge of missing fault labels in the target operating condition, we propose a knowledge transfer approach from source to target operating conditions. A novel Multi-scale and Multi-view Domain Adversarial Network (MMDAN) is designed and experimentally validated using marine engine data. Experimental results demonstrate that the proposed method achieves an average diagnostic accuracy of 96.58%. Furthermore, in partial-set transfer tasks across different operating conditions, MMDAN exhibits superior diagnostic performance compared to other state-of-the-art learning models.

This paper studied the liquefied natural gas maritime inventory routing problem (LNG-MIRP) with the mode of shipping logistics company managed inventory. Based on the relationship between LNG boil off rate, temperature difference inside and outside the cabin, and cargo volume, the LNG boil off functions for each voyage and port loading and unloading process were established. An LNG-MIRP nonlinear stochastic programming model was constructed considering changes of LNG boil off rate with the objective of minimizing the total cost of maritime logistics companies. Then, the model was transformed into a mixed integer linear programming model by employing the two model transformation methods of chance constraint and piecewise linear secant approximation. Taking the LNG project from Yamal to China as an example, the proposed model and its algorithm were validated and analyzed. The results show that considering changes of boil off rate within LNG-MIRP can significantly reduce ship fuel costs and LNG boil off losses and can save the total logistics cost. Sensitivity analysis indicates that an increase in LNG prices will lead to an increase in total logistics costs. The research conclusions can provide useful reference for LNG-MIRP decision-making of maritime logistics companies.