基于切换视线法的欠驱动无人艇鲁棒自适应路径跟踪控制

针对欠驱动无人艇（USV）的航路点路径跟踪问题，提出了一种鲁棒自适应控制方法。设计了一个新的切换型视线（LOS）法制导律，以引导USV始终以最佳LOS圆半径趋向期望路径。与传统制导律相比，切换型LOS法制导律具有运算负载小、收敛速度快的特点。考虑到USV模型的不确定性及环境干扰未知时变性，开发了一种复合神经网络控制器来增强系统鲁棒性。通过在网络输入中增加预测误差信息，以提高网络逼近精度。同时，引入最小学习参数思想优化网络结构，对神经网络的在线自适应参数进行压缩，以减轻网络计算负担，并借助李雅普诺夫理论对系统的稳定性进行了分析。通过对比仿真实验验证了所提出控制策略的有效性。

Switching-line-of-sight-guidance-based Robust Adaptive Path-following Control for Underactuated Unmanned Surface Vehicles

A robust adaptive control method is proposed for the way-point-based path following control of underactuated unmanned surface vehicles (USVs). A switching line-of-sight (LOS) guidance law is designed to keep USV always running to a desired path with the best LOS circle radius. Switching LOS guidance law has the advantages of less computing burden and faster convergence compared to the traditional guidance methods. Considering the system uncertainties of vehicle and the unknown time-varying disturbances, a composite neural network controller is developed to enhance the robustness of the system. The approximation accuracy of network is improved by adding the predicted error information in the network input. The number of online adaptive parameters is reduced, which effectively lightens the computing burden, by using the minimal learning parameter techniques to optimize the network structure. The stability of system is analyzed based on the Lyapunov theory. The simulated results are presented to demonstrate the effectiveness of the proposed control strategy.