As one of the ship energy efficiency optimization measures with the most energy saving and emission reduction potential, ship speed optimization has been recommended by the International Maritime Organization. In ship speed optimization, considering the influence of weather conditions, route segmentation and weather data loading methods significantly affect the reliability of speed optimization results. Therefore, taking the ocean-going container ship as the research object, on the basis of constructing the main engine fuel consumption prediction model and shaft speed prediction model based on machine learning methods, a route segmentation and weather loading-speed optimization iterative algorithm is proposed in this study. Single-objective speed optimization research is then conducted based on the algorithm. The research results show that the proposed algorithm effectively reduces the difference between optimized fuel consumption and actual fuel consumption, and can achieve a fuel-saving rate between 2.1% and 5.2%. This study achieves an accurate and reliable prediction of ship fuel consumption and shaft speed, and solves the strong coupling problem between route segmentation, weather loading, and speed optimization by iterative optimization of ship speed. The proposed algorithm provides strong technical support for ships to achieve the goal of energy saving and emission reduction.

Aiming to investigate the transient four-quadrant hydrodynamics on blades under the circumferential motion of the steering center, an efficient three-dimensional (3D) model with half blades was established in this paper. The 3D model with half blades accounts for the hydrodynamic loss induced by the complicated mechanical structure, predicting the hydrodynamic performance of the high-load cycloidal propeller with high accuracy. On this basis, the open water manoeuvering performance of the cycloidal propellers in fully azimuth angle was simulated by Reynolds-averaged Navier-Stokes (RANS) solver, and the hydrodynamic loads and flow field characteristics were analyzed in detail. The research shows that the variation of azimuth angle φ induces the rapid response of hydrodynamic loads, which directly affects the direction of ship motion. Compared with φ = 0°, the increase in the azimuth angle of the steering center will enhance the negative impact of the inflow velocity on the wake flow field and the internal flow field within the blades. Additionally, the non-uniform flow in the main thrust direction leads to a significant increase in the undesired net lateral force. Our primary findings revealed in present paper should contribute to accurate prediction of hydrodynamic performance for high-load cycloidal propellers during maneuvering process.

Wind and waves are the main factors of environmental loading on ships and offshore structures. Thus, detailed understanding of wind and wave conditions can improve the design and maintenance of these structures. This paper developed a validated long-term wind and wave hindcast database covering the recent 32 years from 1988 to 2019. The spatial distribution of wind and wave characteristics for the whole Xisha Islands’ domain were analyzed. Frequency and directional distributions of wind speeds and significant wave heights were investigated at several locations around typical islands. Extreme value models were used to estimate the wind speed for 100-year return levels, whereas environmental contour approaches were utilized to establish the extreme sea-state parameters for 50- and 100-year return periods. It was found that the Weibull distribution was better fitted to the significant wave heights of the Xuande Atoll’s sites in the open sea, while the exponential Weibull distribution provided a better fit at the Yongle Atoll’s sites where waves are sheltered.

Path following is crucial for enhancing the autonomy of unmanned surface vehicles (USVs) in water monitoring missions. This paper presents an offline reinforcement learning (RL) controller for USVs. The controller employs the soft actor–critic algorithm with a diversified Q-ensemble to optimize the steering control policy using a pre-collected dataset of USV path-following trials. A Markov decision process (MDP) tailored for path following is formulated. The proposed offline RL steering controller, trained on static datasets, demonstrates improved sample efficiency and asymptotic performance due to an expanded ensemble of Q-networks. The accuracy and adaptive learning capabilities of the RL controller are validated through simulations and free-running tests.