基于LS-SVM的船舶动力定位有限时间控制器设计

Finite-time Controller Design Based on LS-SVM for Dynamic Positioning Ships

为解决船舶模型存在的具有任意不确定性特征的船舶动力定位控制问题,基于有限时间Lyapunov理论提出一种非奇异快速终端滑模控制策略(NFTSMC),提高系统的收敛速度和抗干扰能力。针对模型不确定性问题,利用最小二乘支持向量机(LS-SVM)的非线性函数逼近技术进行补偿控制,引入"最小参数"技术,将在线学习参数减少到1个,解决"维数灾难"问题。仿真对比结果表明,提出的控制策略具有较高的优越性。

In order to solve dynamic positioning control problem of marine ships with arbitrary uncertainties, a nonsingular fast terminal sliding mode control(NFTSMC) approach is presented based on the Lyapunov criterion, ensuring that the control system owns faster convergence and disturbance rejection properties. Aiming to the arbitrary uncertainties, the least squares support vector machines(LS-SVM) and "minimal learning parameter" technique are introduced to compensate the uncertainties of ships model, and the number of online learning weights used for nonlinear function approximation are decreased substantially to one, solving the problem of "curse of dimensionality". The simulation comparison results show that the proposed control strategy has higher superiority.