Abstract This paper solves the trajectory tracking problem for underactuated autonomous surface vessels in the presence of uncertainty dynamics and unavailable velocities. A robust controller is proposed by employing a neural network, command filtered backstepping method, and adaptive control method. Moreover, all tracking errors are guaranteed to be uniformly ultimately bounded on the basis of the Lyapunov Theorem. The findings of the study are summarized as follows: (i) The uncertainty dynamics of the control system are estimated and approximated by the neural network and state predictors, which allows the designed controller to be easily applied in practice. (ii) To accurately acquire the velocities of the control system, a neuro-adaptive observer is proposed to obtain the unavailable velocities. (iii) A filtered compensation loop is built to decrease filtered signal error, which is caused by the second-order filter. Finally, simulations are performed to verify the robustness and effective tracking performance of the proposed control scheme in consideration of external disturbances.

中文翻译：

---

具有不确定动态和不可用速度的欠驱动自主水面舰艇的鲁棒轨迹跟踪控制

摘要 本文解决了存在不确定动力学和不可用速度的欠驱动自主水面舰艇的轨迹跟踪问题。通过采用神经网络、命令过滤反步法和自适应控制方法，提出了一种鲁棒控制器。此外，根据李雅普诺夫定理，保证所有跟踪误差最终一致有界。研究结果总结如下： (i) 控制系统的不确定性动态由神经网络和状态预测器估计和近似，这使得设计的控制器易于实际应用。(ii) 为了准确获取控制系统的速度，提出了一种神经自适应观察器来获取不可用的速度。(iii) 建立滤波补偿回路以减少由二阶滤波器引起的滤波信号误差。最后，进行仿真以验证所提出的控制方案在考虑外部干扰的情况下的鲁棒性和有效跟踪性能。