The paper proposes a new robust adaptive control method for a two-link robot manipulator to enhance high tracking control performance despite the presence of dynamic uncertainties and unknown disturbances. The paper examines a Non-Singular Fast Time Sliding Mode (NFTSM) controller based on Wavelet Neural Network (WNN). The Wavelet Network is employed to approximate the upper bound of uncertainties and disturbances. In addition, a compensation term is added to the NFTSM control to attenuate the effect of uncertainties, including unavoidable approximation errors and unknown disturbances. Therefore, to ensure high tracking accuracy, chattering phenomenon reduction, and fast response against approximation errors and uncertainties. The parameters of the controller are tuned online by an adaptive learning algorithm, and online adaptive control laws are determined by the Lyapunov stability theorem. Due to these techniques, the suggested control drives the system to the desired performance where tracking errors converge to zero within a finite time. Simulations performed on a two-link robotic arm demonstrate a higher performance of the proposed adaptive robust NFTSMC based on WNN methodology compared to some advanced control strategies.

本文提出了一种新的鲁棒自适应控制方法，该方法适用于两链机器人，尽管存在动态不确定性和未知干扰，但仍能提高高跟踪控制性能。本文研究了基于小波神经网络（WNN）的非奇异快速滑模（NFTSM）控制器。小波网络用于近似不确定性和扰动的上限。另外，将补偿项添加到NFTSM控件以减弱不确定性的影响，其中包括不可避免的近似误差和未知干扰。因此，要确保较高的跟踪精度，减少抖动现象以及对近似误差和不确定性的快速响应。控制器的参数通过自适应学习算法在线调整，在线自适应控制定律由Lyapunov稳定性定理确定。由于这些技术，建议的控制将系统驱动到所需的性能，其中跟踪误差在有限时间内收敛为零。与某些先进的控制策略相比，在双链接机械臂上进行的仿真表明，基于WNN方法的自适应鲁棒NFTSMC自适应系统性能更高。