Xiaoyong Shang

Autonomous path planning and collision avoidance (CA) in complex maritime environments are critical challenges for maritime autonomous surface ships (MASSs). This article presents a novel hierarchical motion planning framework that integrates global and local path-planning algorithms. For global path planning in static-obstacle environments, a hybrid approach combining A*, Bezier curves, and particle swarm optimization (PSO) is proposed to enhance path search efficiency and ensure smooth trajectory optimization. For local path planning and CA with dynamic vessels, a multiship CA algorithm based on velocity prediction potential fields (VPPFs) is introduced. A new collision-risk identification model is developed to ensure compliance with COLREGs and enable collaborative CA in complex multiship scenarios. The framework also addresses environments with both dynamic and static obstacles through a dynamic path planner that integrates global and local planning. A key innovation is the dynamic CA decision making mechanism with alteration of course and/or speed (ACS), which provides robust real-time navigation strategies. Validation through simulations, real-ship model tests, and automatic identification system (AIS)-data analysis demonstrates the framework’s robustness and adaptability for MASS navigation. Compared to existing methods, the proposed approach significantly reduces collision-avoidance path length while maintaining safety, real-time performance, and COLREGs compliance. This work advances autonomous maritime navigation by addressing complex scenarios with a unified framework that balances efficiency, safety, and regulatory adherence.

Research on unmanned surface vessels (USVs) has evolved significantly in recent decades. In particular, intelligent navigation technology has progressed from theoretical concepts to practical applications. As USV research in ocean engineering advances, there is an increasing demand for enhanced performance in intelligent guidance strategy and path-following control systems. This manuscript proposes future development directions for USVs by providing an overview of relevant standards for the intelligence level of these vessels and describing the current status of USV engineering practices. Based on practical ocean engineering requirements, safety considerations, and energy efficiency demands, this paper summarizes the current research status, future research challenges, and potential solutions for USV intelligent guidance and path-following control algorithms from the perspective of large ship intelligence. This manuscript provides a valuable reference for academic researchers and practitioners aiming to identify and position future development directions.

This manuscript investigates a novel scheme designed for path planning and path following of the USV system to perform dynamic berthing tasks in the presence of structural uncertainties and simulated marine environments. The proposed cooperative scheme consists of the dynamic artificial potential field ship (DPFS) guidance principle and a two-dimensional trigger iterative learning control algorithm. The DPFS guidance principle employs a novel gravitational potential field to plan desired yaw angles and a spiral repulsive potential field to generate smoother reference paths. This principle is developed to plan the desired path for the unmanned dinghy to perform dynamic berthing tasks with the maintenance ship. To ensure the USV converges to the desired path with superior control accuracy, an improved iterative learning control algorithm is designed with a two-dimensional event-triggering mechanism. The event-triggering mechanism in the time domain reduces actuator wear by decreasing the transmission frequency of control input. As well as, in the iteration domain allows for the presetting of control error based on ocean engineering tasks, and the number of iterations can self-adjust accordingly. Utilizing the Lyapunov theorem, the proposed control scheme guarantees uniform convergence of full state tracking error over the iteration domain. The superiority and effectiveness of the proposed strategy are validated through numerical experiments conducted on the MATLAB platform under various external disturbances.