基于神经网络的多机械臂固定时间同步控制

Fixed-time synchronization control of multi-manipulator based on Neural Network

摘要:

针对动力学模型未知的多机械臂系统，提出了一种基于神经网络的固定时间终端滑模的位置同步控制器。首先结合相邻交叉耦合同步控制策略，设计固定时间终端滑模面与控制器，保证系统的跟踪误差与同步误差在固定时间内收敛，且收敛时间上界与初始状态无关。其次，设计RBF神经网络权值更新律估计系统多机械臂未知非线性动力学模型，该方法无需对系统模型参数的先验知识。利用Lyapunov函数证明系统的固定时间收敛性与稳定性。最后，仿真结果验证了所提方法的有效性。

Abstract:

In this paper, a fixed-time sliding mode position synchronization controller based on neural network is proposed for multi-manipulator system with unknown dynamics model. Firstly, a fixed-time terminal sliding mode surface and controller are designed based on the adjacent cross-coupled synchronization control strategy to ensure that the tracking error and synchronization error converge in a fixed-time, and the upper bound of the convergence time is independent of the initial state. Secondly, the weight updating law of RBF neural network is designed to estimate the unknown nonlinear dynamic model of the system. This method does not need prior knowledge of the parameters of the system model. The Lyapunov function is used to prove the fixed-time convergence and stability of the system. Finally, simulation results verify the effectiveness of the proposed method.