欢迎访问中国科学院大学学报,今天是

中国科学院大学学报 ›› 2011, Vol. 28 ›› Issue (4): 514-521.DOI: 10.7523/j.issn.2095-6134.2011.4.013

• 论文 • 上一篇    下一篇

漂浮基空间机器人的基于模糊神经网络的自适应补偿控制

张文辉1, 齐乃明1, 马静2, 肖阿阳1   

  1. 1. 哈尔滨工业大学航天学院, 哈尔滨 150001;
    2. 东北农业大学工程学院, 哈尔滨 150001
  • 收稿日期:2010-08-17 修回日期:2010-09-28 发布日期:2011-07-15
  • 基金资助:

    中国航天科技集团创新基金(CAST09C01)资助 

Adaptive compensation control of free-floating space robot based on fuzzy neural network

ZHANG Wen-Hui1, QI Nai-Ming1, Ma Jing2, XIAO A-Yang1   

  1. 1. School of Aerospace, Harbin Institute of Technology, Harbin 150001, China;
    2. Department of Engineering, Northeast Agriculture University, Harbin 150001, China
  • Received:2010-08-17 Revised:2010-09-28 Published:2011-07-15

摘要:

针对自由漂浮状态的空间机器人模型不确定性及其动力传动机构的摩擦死区非线性,将一种自适应模糊小脑模型关联控制(FCMAC)补偿策略用于轨迹跟踪及补偿问题.利用模糊神经网络并引入GL矩阵及其乘法算子"."分别对执行机构中的摩擦死区及系统模型不确定部分进行自适应补偿,其补偿误差及外界扰动通过滑模控制器来消除.基于Lyapunov理论证明了闭环系统跟踪误差的有界性.仿真表明控制器可以达到较高精度,且能满足实时性要求.

关键词: 模糊神经网络, GL矩阵, 空间机器人, 自适应控制

Abstract:

Considering trajectory tracking of free-floating space robot with uncertainties and friction blind section non-linearity, we propose an adaptive fuzzy CMAC compensation control algorithm. The control scheme uses fuzzy neural network to establish modeling online, and imports GL matrix and multiplication operator "." into neural network to distinguish parameters of system and friction blind section non-linearity. The control scheme can guarantee the stability of closed loop system and the asymptotic convergence of tracking errors. Neural network approach errors and outside disturbance can be eliminated by sliding model controller. Based on a standard Lyapunov theorem, we prove that all signals in the closed-loop are bounded. The simulation results show that the controller can achieve high control precision and meet the requirement of real time.

Key words: fuzzy neural network, GL matrix, space robot, adaptive control

中图分类号: