不确定性机器人的线性参数神经网络跟踪控制

考虑机器人建模误差不确定性,首先利用反馈控制技术,把基于拉个朗日方程的多关节机器人动力学模型转化成一线性状态方程。然后基于此状态方程,运用线性参数神经网络设计出鲁棒神经网络控制器,并且用李雅普诺夫函数思想证明了这种控制器能够使得机器人系统满足指数收敛的特性。

On Tracking Control for Uncertain Robot Based on Neural Network

Considering the uncertainty of robot modeling errors,the multi-joint robot dynamics model on the basis of Lagrange equation is first transformed into a linear state equation via feedback control technique,then,corresponding to the linear state equation,by making use of linear-parameter-neural-networks(LPNN),the robust controller is designed,which can be proved by taking the Lyapunov function methods that it makes the robot systems satisfy the exponential convergence.