The paper addresses an adaptive robust position control for tracking control of a manipulator under the presence of the uncertainties, such as variant payload, modeling error, friction, and external disturbance. The proposed control uses radial basis function neural networks (RBFNN)s to approximate and cancel the uncertainties. The nonsingular fast terminal sliding mode control (NFTSMC) of the proposed control is developed to guarantees a finite-time convergence and to solve the singular issue of the terminal sliding mode control. Moreover, the learning laws are derived from the Lyapunov approach to ensure the stability and robustness of the whole system. The proposed control is compared with other controllers through both simulations and experiments on a 3-DOF manipulator to exhibit its efficiency with the variant payload and the uncertainties.

本文提出了一种自适应鲁棒位置控制，用于在不确定性（例如变量有效载荷，建模误差，摩擦和外部干扰）存在的情况下对机械手进行跟踪控制。所提出的控制使用径向基函数神经网络（RBFNN）来近似和消除不确定性。提出的控制系统的非奇异快速终端滑模控制（NFTSMC）是为了保证有限时间收敛并解决终端滑模控制的奇异问题而开发的。此外，学习定律源自李雅普诺夫方法，以确保整个系统的稳定性和鲁棒性。通过在3-DOF机械手上进行仿真和实验，将拟议的控制与其他控制器进行比较，以显示其有效载荷和不确定性的效率。