Guibing Zhu, Xinxin Peng, Yong Ma, Songlin Hu
Aiming at the problem of obstacle avoidance path tracking control for unmanned surface vessels (USVs) affected by time-varying disturbances and dynamic uncertainties, this work proposes a planning-guidance-control integrated design framework. In the obstacle avoidance layer, pioneers use visual projection technology (VPT) with an obstacle avoidance path optimization mechanism (OAPOM) to generate safe, maneuverability-compliant paths. In the guidance layer, to handle the underactuated problem and compensation problem of sideslip angle, a novel finite-time integral line-of-sight (FTILOS) scheme with time-varying sideslip angle estimator is developed. At the control layer, a virtual parameter adaptive neural network controller based on FTILOS is developed to learn online and suppress compound uncertainties. Compared with the existing work, this work innovatively proposes a three-module design framework, i.e., the OAPOM-based obstacle-avoidance path seamlessly connects with FTILOS guidance, while the guidance layer provides ideal heading and speed signals for the control layer, forming an integrated path planning–guidance–control design architecture. Lyapunov theory proves that all signals in a closed-loop path following obstacle avoidance control system are uniformly bounded. High-fidelity simulation and water experiments confirm robustness, path feasibility, and control scheme practicality.