Maritime rescue missions require efficient and reliable path planning for unmanned surface vehicles (USVs). However, these missions are challenged by the limited sensing capabilities of USVs operating in vast and uncertain environments with randomly distributed obstacles. This study addresses the issues of low path planning efficiency and poor robustness resulting from restricted perception range. To tackle these challenges, a novel local observation-based path planning approach is proposed for USVs in maritime rescue missions.

The proposed approach integrates three key methodological innovations. First, the soft actor-critic (SAC) algorithm is employed with a reward function tailored to local observation, which rewards efficient goal-reaching and penalizes obstacle collisions. This design helps balance exploration and exploitation in uncertain environments. Second, a feature-enhanced soft actor-critic (FESAC) algorithm is introduced to improve training efficiency and model robustness. It extracts key environmental features and employs a randomized training environment with strategically placed obstacles to enhance sampling efficiency. During training, obstacle positions, USV starting points, and goals are randomly reset across episodes, encouraging the model to learn generalizable navigation strategies instead of memorizing specific scenarios. Third, an adaptive waypoint planning algorithm is developed based on local perception domains to effectively coordinate local obstacle avoidance with global goal-reaching behavior. Waypoints are dynamically selected within the USV's perception radius using a weighted objective function that balances proximity to the goal and distance from obstacles. This decomposes the complex global path planning task into a series of manageable local planning problems.

Comprehensive simulation experiments validate the effectiveness of the proposed approach. In feature-rich environments with randomly distributed obstacles, the method achieves a success rate exceeding 98%, significantly outperforming traditional methods. In simulated maritime rescue missions over 1 000 m×1 000 m areas with 20−50 randomly placed obstacles, the method maintains a task completion rate exceeding 93% under appropriate parameter configurations. The simulation results also reveal a notable trade-off between path safety and efficiency: increasing the obstacle avoidance weight $w_2$ yields safer but longer paths, whereas increasing the goal-reaching weight $w_1$ results in shorter paths at the cost of higher collision risk. Depending on different task requirements, optimal performance metrics can be obtained through proper parameter tuning. Comparative analysis shows that the FESAC algorithm converges significantly faster than standard SAC in complex environments, demonstrating enhanced learning efficiency.

The proposed local observation-based path planning method effectively addresses the challenges posed by limited perception in maritime rescue scenarios, exhibiting strong robustness and adaptability to uncertain environments. By decomposing complex global planning tasks into manageable local subtasks and enhancing feature extraction capabilities, the method provides a practical solution for real-world USV operations where complete environmental information is unavailable. This work provides valuable technical insights for the practical application of reinforcement learning algorithms in actual engineering scenarios.

This study addresses the critical challenge of precise waypoint tracking for an underactuated unmanned surface vehicles (USV) in complex marine environments. The primary objective is to integrate path planning with guidance and control to enhance tracking accuracy, ensure system reliability, and optimize overall performance. The research aims to provide a robust and efficient solution for autonomous USVs, which are essential for various maritime missions, including environmental monitoring, resource exploration, search-and-rescue operations, and military applications.

This study proposes a control framework named NRRT−FTC (non-uniform rational B-spline-based rapidly exploring random tree fault-tolerant control), designed to improve path planning and waypoint tracking for USVs. First, the framework utilizes an improved rapidly exploring random tree (RRT) algorithm to efficiently select waypoints and remove redundant points from the path. The waypoints are then fitted using non-uniform rational B-splines (NURBS), a powerful technique that ensures the generation of a smooth and accurate path passing through all the waypoints. This NURBS-based approach not only optimizes the smoothness of the generated path but also enhances the controllability and feasibility of the trajectory, making it suitable for practical applications in marine environments. Furthermore, the NRRT−FTC framework incorporates a fault-tolerant control strategy based on finite-time observers. This strategy plays a critical role in managing environmental disturbances, system uncertainties, and actuator failures. By performing real-time state estimation and adjusting the control inputs accordingly, the framework ensures that the USV can continue to track the planned path accurately, even under failure conditions. This design improves system robustness and ensures reliable performance under diverse operational scenarios.

Simulation experiments were conducted to validate the effectiveness of the NRRT−FTC framework. The experimental results demonstrated significant improvements in both path planning and tracking control compared to other methods. Specifically, the control error was reduced by 31.78%, control efficiency increased by 16.65%, and task execution efficiency was improved by 36.17%. The NRRT−FTC framework also demonstrated strong resilience to environmental disturbances and actuator faults, ensuring precise waypoint tracking even under adverse conditions including wind, waves, and sensor noise.

The NRRT−FTC framework significantly improves the waypoint tracking accuracy of underactuated USVs, thereby enhancing system robustness and reliability. By integrating path planning and guidance control, the framework provides a practical solution for precise navigation in marine missions, particularly in challenging and unpredictable environments. The findings contribute to USV autonomy, providing a robust theoretical foundation for future advancements in autonomous vehicle technology for maritime applications. The proposed approach has significant potential for real-world applications, including autonomous navigation in complex marine environments and beyond.

To overcome the dynamic constraints faced by an unmanned surface vehicle (USV) in sea area coverage tasks, a coverage path planning scheme is constructed that considers the motion characteristics of USVs.

For obstacle-free environments, polygon region coverage path planning is used to generate coverage paths. For obstructed or restricted environments, trapezoidal decomposition is adopted to decompose the target area into several sub-regions, and a genetic algorithm (GA) is used to optimize the link path between each sub-region. Based on the predicted trajectory set of the USV maneuvering motion model, the turning path and its corresponding reference propeller speed are optimized, and the full coverage of the target sea area is achieved.

The simulation results show that the proposed method can achieve 100% path coverage in sea areas containing obstacles, with each turning path satisfying the maneuvering motion characteristics of the USV.

The full-coverage path planning scheme proposed herein can generate a path that satisfies the maneuvering constraints of the USV, giving it obvious practical value in the fields of intelligent fishery, marine exploration, and marine communication relay.

This paper proposes a speed and power allocation strategy collaborative optimization method for hydrogen fuel cell ships based on particle swarm optimization (PSO) in order to solve the problem of the energy efficiency improvement of such ships being limited by their difficulty in reaching optimal speed under complex sailing environments.

The spatial grid clustering analysis of meteorological environment data is carried out by K-means and used as the basis for route segmentation. A speed-hydrogen consumption model of a fuel cell ship is constructed by ship navigation resistance analysis and the equivalent hydrogen consumption idea. The optimization method of inheritance-link between voyage segments is designed by taking acceleration as the optimization parameter, then PSO is used to optimize the ship's speed and fuel cell output power.

The results show that the proposed speed and power allocation strategy collaborative optimization method reduces hydrogen consumption by 3.85% and 1.99% respectively compared with the original speed and traditional speed segmentation optimization method.

The proposed method can effectively improve the sailing energy efficiency of short-range ships and resolve the defect of the speed ladder allocation of traditional segmental optimization, providing useful references for the popularization and application of hydrogen fuel cell ships.

In view of the effects of life decay on fuel cell efficiency and power system performance, an adaptive equivalent hydrogen consumption minimization strategy is proposed that considers the life decay of fuel cells.

Based on the relationship between fuel cell efficiency and degradation rate, a fuel cell degradation factor is introduced to design an energy management objective function that considers efficiency degradation. The influence factors of fuel cell degradation are then analyzed, and a power distribution optimization system that considers both economy and durability is constructed. Finally, the optimization performance of the proposed strategy in the whole life cycle of the fuel cell is verified by comparison experiments under three typical scenarios: slight degradation, semi-degradation and severe degradation.

Compared with the original strategy, the proposed strategy can reduce fuel cell degradation by up to 12.4% and ship voyage cost by 4.97% during the whole life cycle of the fuel cell.

This method can effectively improve the economy and durability of marine hybrid power systems.

This study explores the limited thrust/torque of a remotely operated vehicle (ROV) by addressing the difficulty of achieving precise control in prescribed-performance 3D trajectory tracking. Considering unknown factors such as system uncertainty and underwater environmental disturbances, an accurate tracking control scheme is proposed on the basis of a finite-time extended state observer and prescribed-performance transformation to ensure the rapid stabilization of trajectory tracking errors.

First, a compensation system is designed to eliminate the thruster input saturation constraint. Second, a finite-time extended state observer is designed to perform lumped observation and compensation for external disturbances and unknown system dynamics. Furthermore, based on the prescribed-performance function and error conversion function, the tracking error limited by the prescribed-performance is transformed into an unrestricted tracking error; an integral sliding mode is constructed; and the fast power reaching law and boundary layer are utilized to reduce the buffeting of the actuator. Finally, the Lyapunov theory is used to prove the overall stability of the proposed algorithm.

The simulation results verify the effectiveness and superiority of the designed control method.

This study can provide a new solution for the accurate prescribed-performance control of the trajectory tracking of an ROV with constrained thrust under lumped disturbance.

Aiming at the accurate posture stabilization problem of an under-actuated unmanned surface vehicle (USV) in GPS-denied environments, a monocular visual servo stabilization control scheme is proposed based on homography.

By virtue of the homography decomposition technique, posture errors with an unknown scale factor are directly reconstructed from current and desired images, which thoroughly removes the calibration of extrinsic camera parameters and priori information on visual targets; with respect to the under-actuation constraint, a periodic function to persistently excite the yaw angle is incorporated into the continuous time-variant output feedback controller, allowing the USV to be stabilized in the absence of image depth, movement velocities and model parameters.

Under the framework of the Lyapunov theory, the closed-loop visual servo system of the USV is rigorously proven to be asymptotically stable by Barbalat lemma.

By installing an onboard monocular camera, USV posture errors can be precisely stabilized with the aid of the proposed visual servo strategy, providing significant technique support for practical applications including docking, berthing, dynamic positioning, etc.

This paper presents a non-singular terminal sliding mode track control method based on a finite-time disturbance observer to solve the problem of accurately tracking and controlling an the 3D trajectory of an unmanned underwater vehicle under complex external disturbances.

A non-singular terminal sliding mode track controller is designed to ensure that the tracking error converges to zero accurately within a limited time. A finite-time disturbance observer is designed to improve the anti-jamming ability of the system under external multidimensional time-varying disturbances.

The Lyapunov function is used to prove that the designed control strategy can remain stable for a limited time. MATLAB is used for the simulation experiment, and a comparison with the backstepping sliding mode control method under step disturbance shows that the method presented in this paper achieves accurate trajectory tracking.

The results of this paper can provide a solution for accurately tracking the 3D trajectories of unmanned underwater vehicles.