欠驱动船舶自适应神经网络有限时间跟踪控制

Adaptive neural finite-time tracking control of underactuated marine surface vessel

针对受动态不确定性和外界未知干扰影响下欠驱动水面船舶的轨迹跟踪控制问题,设计一种有限时间轨迹跟踪控制方案。采用神经网络重构船舶的动态不确定性,通过引入最小学习参数降低计算复杂度,设计自适应律逼近由不确定项和外界干扰组合而成的复合扰动的上界,并基于此设计一种基于最小学习参数的欠驱动船舶自适应神经网络有限时间轨迹跟踪控制方案。通过严格的理论分析后得出,该有限时间轨迹跟踪控制方案能够使闭环系统的所有信号都趋于有界,欠驱动船舶的位姿误差和速度误差都在有限时间内收敛到一个集合。仿真和比较验证了本研究所提出的有限时间控制方案的有效性。本研究中的有限时间控制方案不仅提高了船舶的瞬态性能和稳态性能,且控制器结构简单,更容易应用在工程中。

This research provided a finite-time trajectory tracking control scheme for underactuated marine surface vessels which were influenced by dynamic uncertainties and unknown disturbances.In the scheme,the internal uncertainties were approximated by the neural networks.To decrease the computing complexity,the minimum learning parameters was applied.The compound disturbance composed by dynamic uncertainties and unknown disturbances was estimated by designing an adaptive law.An adaptive neural finite-time trajectory tracking control of underactuated marine surface vessels based on minimum learning parameters was designed.Rigorous theoretical analysis were provided to prove that the finite-time trajectory tracking control scheme could make all signals of the closed-loop system tended to be bounded.And it could ensure that the pose error and velocity error of the underactuated marine surface vessels could converge to a set in a finite time near the origin.The effectiveness and preponderance of the proposed control scheme were demonstrated by simulations and comparative results.The finite-time control scheme in this study not only improved the transient performance and steady-state performance of the vessel,but also had a simple controller structure and was easier to apply in engineering.