A high-precision iterative control strategy based on automatic identification system (AIS) trajectory reproduction tasks is designed to address the low tracking accuracy, system model uncertainties, and disturbances from time-varying marine environments encountered by underactuated ships.

First, a guidance method driven by AIS data is proposed, integrating AIS data with the dynamic virtual ship (DVS) guidance principle. Second, robust neural damping technology and dynamic surface control (DSC) are introduced to approximate model uncertainties and complex derivatives of virtual control laws, thus avoiding the issue of “computational explosion”. Furthermore, an iterative learning control strategy is incorporated based on traditional robust control methods to complete the final controller design. Finally, the semi-global uniformly ultimately bounded (SGUUB) stability of the proposed controller is demonstrated using the Lyapunov theorem. A simulation experiment is conducted to validate its performance under time-varying marine environmental disturbances.

The results indicate that the path following of underactuated ships under this control strategy exhibits high precision and robustness, with average control accuracy maintained within 0.5 meters.

The research results are of great significance for safe ship navigation and trajectory reproduction in complex waters.

To meet the demands for high-precision path tracking and communication resource optimization in unmanned surface vehicle (USV) formations navigating in non-cooperative target tracking scenarios, this paper proposes a novel prescribed performance formation control algorithm integrated with a dynamic event-triggered mechanism (DETM).

First, an adaptive guidance switching framework based on a virtual leader-follower structure is developed. During waypoint-based synchronous cruising, the formation structure is maintained by fixing relative distances and bearings. When facing a non-cooperative target vessel, decoupling conditions for formation reference signals are introduced, along with a velocity correction mechanism and an adaptive law to estimate the target vessel’s motion characteristics, thereby ensuring autonomous reconfiguration capability. Second, to overcome the strict sign constraints of boundary functions in traditional prescribed performance control (PPC) during initial error stabilization, a novel set of performance boundary functions is proposed to confine arbitrary initial errors within designated boundaries. Through a mapping transformation, the tracking error constraint problem is transformed into a stabilization problem of new variables, thereby ensuring that the formation’s position and heading errors remain strictly within the prescribed performance limits. Additionally, a dual-threshold regulated DETM is introduced to effectively reduce communication load between controllers and actuators. Unlike static mechanisms, the dynamic mechanism enables online adaptive updates of time-varying thresholds based on system states, while theoretically eliminating Zeno behavior. Meanwhile, radial basis function neural networks (RBF-NNs) are employed to approximate inherent nonlinearities in the model and unknown external disturbances, and are combined with dynamic surface control (DSC) to avoid repeated differentiation of virtual control laws. Based on Lyapunov stability theory, the semi-globally uniformly ultimately bounded (SGUUB) stability of the closed-loop system is proven. Finally, two simulation experiments quantitatively validate the algorithm.

The results show that the proposed algorithm achieves high precision, strong robustness, and low communication load during path tracking, with an average position error below 0.5 meters and a heading error within 5 degrees. Compared to static event triggering, the DETM further reduces control update frequency by 8.79% for surge moment and 11.81% for yaw moment while maintaining tracking stability. The framework successfully demonstrates complete processes including cooperative cruising, autonomous target eviction, dynamic decoupling, and reconfiguration, achieving optimized control accuracy and efficient use of communication resources.

The proposed dynamic event-triggered prescribed performance formation control framework integrates velocity correction with adaptive compensation, enabling high-accuracy tracking and enhanced communication efficiency in non-cooperative scenarios. This approach provides new theoretical and technical foundations for non-cooperative control and adaptive reorganization in unmanned surface vehicle formation systems.