摘要
针对更具工程价值的任务空间内的自由漂浮状态空间机器人模型不确定性末端控制问题,提出了一种自适应神经网络控制策略用于机器人末端控制。通过神经网络在线建模来逼近系统中的非线性模型,神经网络的逼近误差及外界有界扰动通过鲁棒控制器来消除,采用引入GL矩阵及其乘法算子"."来直接辨识的各部分系统参数。该控制策略既不需要逆动态模型的估计值,同时也避免了求雅克比矩阵的逆,降低了计算量。基于李亚普诺夫理论证明了整个闭环系统全局渐近稳定。仿真结果表明了这种神经网络控制器对于任务空间的空间机器人末端控制在达到较高的精度的同时,能够满足实时性要求,具有重要的工程应用价值。
Considering free-floating space robot in task space is a neural network adaptive control algorithm is proposed for trajectory tracking in this paper.The control scheme uses neural networks to approach the non-linearity model by establishing online model,and neural network approach errors and outside bounded disturbance can be eliminated by a robust controller.Parameters of system are distinguished by import GL matrix and multiplication Operator ＂·＂ into neural networks.The controller has no need of estimate value of inverse dynamic model and avoids the Jacobian matrix inverse operation.The control scheme can guarantee the stability of closed loop system and the asymptotic convergence of tracking errors based on the lyapunov theory.The simulation results show that the presented controller is effective in control accuracy and can satisfy request for real time,and has important value for engineering application.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2011年第6期1312-1317,共6页
Journal of Astronautics
基金
中国航天科技集团创新基金资助项目(CAST09C01)
关键词
神经网络
GL矩阵
任务空间机器人
自适应控制
乘法算子
Neural network
GL matrix
Space robot in task space
Adaptive control
Multiplication operator
