摘要
针对不确定机器人轨迹跟踪问题 ,提出了一种基于神经网络的自适应鲁棒控制 .该控制方案由一个PD反馈和一个神经动态补偿器组成 ,其特点是不需要系统不确定性上界的先验知识 ,而且避免了求解惯性矩阵逆 .通过利用一个RBF神经网络自适应学习系统不确定性的未知上界 ,从而可以有效克服系统不确定性的影响 ,保证机器人系统的输出跟踪误差渐近收敛于 0 .
An adaptive neural robust controller is presented for trajectory tracking of uncertain robot manipulators. This control scheme consists of a PD feedback plus a neural network-based dynamic compensation. The key feature of scheme is that it does not require any a priori knowledge on the upper bound of system uncertainties. Moreover, it needn't compute the inverse of inertia matrix. An RBF neural network is used to learn the unknown upper bound of system uncertainties. It is shown that the proposed control scheme can effectively eliminate the effects of system uncertainties, and guarantees the asymptotic convergence of tracking error.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2000年第6期924-928,共5页
Control Theory & Applications
基金
supported by the National Defence Research Fundation of Science and Technology (99J16.6.1BQ0214)
关键词
神经网络
机器人
不确定性
鲁棒控制
neural network
robot manipulators
uncertainty
robust control
