基于DSC和MLP的欠驱动船舶自适应滑模轨迹跟踪控制

针对存在船舶动态不确定和外界干扰的欠驱动水面船舶轨迹跟踪控制问题,提出一种基于动态面控制(dynamic surface control,DSC)和最小学习参数法(minimal learning parameter,MLP)的自适应滑模控制方法。在控制律的设计过程中,为实现位置跟踪误差的收敛,利用反步法设计船舶前向速度和横漂速度的虚拟控制律镇定轨迹跟踪误差;引入DSC技术,用于消除反步法对虚拟控制求导引起的“微分爆炸”问题;另外,采用MLP技术,以单参数在线学习代替所有权值在线学习,减少控制器的计算量,避免出现“维数灾难”问题,并结合带有“σ-修正”的自适应律,防止参数漂移,易于工程的实现。稳定性分析证明了所设计控制律可以保证轨迹跟踪船舶闭环系统中轨迹跟踪误差信号一致最终有界,仿真结果验证了所设计控制器的有效性。

Adaptive sliding mode trajectory tracking control of underactuatedship based on DSC-MLP

For the trajectory tracking control problem of underactuated ships with parameter uncertainties and external disturbances, an adaptive sliding mode control strategy based on dynamic surface control (DSC) and minimal learning parameter (MLP) is proposed. In order to achieve convergence of position tracking error, the backstepping method is employed to design a virtual control law for surge and sway velocities of ships to stabilize position errors in the controller design. The DSC control technique is introduced to eliminate the “differential explosion” problem of derivation term of virtual control vector of the backstepping method. The MLP technique is employed to reduce the computation burden of controller and avoid the “the curse of dimensionality” problem, which substitutes single parameter online learning for all weights online learning. An adaptive law incorporated with σ-modification is proposed to prevent parameters from drifting. By the stability analysis, it is proved that the proposed control law can guarantee the uniformly ultimate boundedness of trajectory tracking errors in the closed-loop trajectory tracking system of ship. Simulation results validate the effectiveness and the performance of the proposed scheme.