Abstract
Path following is a fundamental capability for autonomous surface vessels (ASVs). A typical path-following algorithm comprises two main modules: guidance and control. In the guidance domain, the vector field (VF) approach has been widely adopted due to its demonstrated superiority over alternative guidance strategies. Consequently, it has been successfully applied to a range of unmanned systems, including ASVs, airships, and quadrotors. However, most existing VF guidance laws are constrained to simple geometric paths, such as straight lines and circular orbits. Therefore, their practical applicability is limited in real-world engineering scenarios. To address this limitation, this paper introduces a novel, continuously differentiable VF capable of handling general curved paths, thereby significantly broadening the application scope of the VF methodology. In terms of control, a major challenge for ASVs lies in achieving long-endurance operation while maintaining robustness against uncertainties. The finite-time uncertainty observer (FTUO)-based adaptive dynamic programming (ADP) approach has been shown to be effective in addressing this challenge. Nevertheless, common FTUO-based ADP methods often suffer from drawbacks such as system chattering, observer peaking, and asymptotic convergence. In response to this situation, this paper proposes a modified FTUO-based ADP method for ASVs. This approach eliminates undesirable effects, i.e., chattering and peaking, and further saves control energy. Both simulation and experimental results verify the effectiveness and advantages of the proposed path-following algorithm. Note to Practitioners-This study tackles a key challenge in marine engineering: enabling the ASV to follow a curved path smoothly and energy efficiently in real ocean conditions. Most existing VF guidance methods are effective for simple paths but exhibit degraded performance for curved paths. This research proposes a smooth VF-based guidance strategy that supports general curved paths. An ADP-based control policy with guaranteed finite-time convergence is developed to reduce energy consumption and achieve long-endurance operation. An FTUO is also built to handle ocean disturbances without inducing chattering, simplifying real-world implementation. Future work will generalize the proposed control framework to multiple ASVs. Compared to single-vessel application scenarios, cooperative control of multiple ASVs introduces greater complexity, particularly in handling collision avoidance, network connectivity, and communication delays. To address these issues, we plan to exploit a barrier function to modify the VF guidance module and the cost function to achieve safe cooperative motion. In addition, we will integrate the Lyapunov-Krasovskii function with the ADP algorithm to ensure robustness to communication delays.