具有时滞的柔性关节多机械臂协同自适应位置/力控制  

Adaptive position/force control for coordinated multiple flexible-joint manipulators with time delay

作　　者：李树荣[1] 尹怀强 

机构地区：[1]中国石油大学(华东)信息与控制工程学院,山东青岛266580

出　　处：《控制理论与应用》2017年第9期1208-1214,共7页Control Theory & Applications

基　　金：国家自然科学基金项目(61573378)资助~~

摘　　要：由于关节机械臂长期运行后,齿轮间隙扩大产生的时间滞后将使得系统跟踪性能降低.针对此问题,本文提出了一种自适应位置/力控制策略来保证闭环系统稳定性以及位置/力跟踪性能.首先,对多机械臂和物体系统进行任务空间动力学建模.随后,利用Pade理论将时间滞后近似为二阶有理分式.同时,利用神经网络自适应算法克服模型建模误差对系统稳定性的影响,利用同时包含位置误差和力误差的线性滑模项,设计位置/力控制器.通过李雅普诺夫稳定性理论,证明控制策略能实现位置误差和内力误差的渐近收敛.最后,仿真验证证明所设计控制策略的有效性.The gear clearance will expand due to the long running joint manipulators. Time delay results from the clearance, which also has influence on system tracking performance. To solve this problem, an adaptive position/force control strategy is proposed to guarantee the stability of the closed-loop system and tracking performance of the position/force.Firstly, the task space dynamic model for the system is given in this paper. Then the time delay is approximated as a second-order rational fraction by Pade theorem. Secondly, consider the model uncertainties, by utilizing a sliding mode consisting of position error and force error, a neural network adaptive controller is proposed. The control strategy can drive the closed-loop system states to reach the adopted sliding model surface and then both position and force errors can be guaranteed to converge zero. Finally, simulation results are presented to validate the proposed method.

关 键 词：协同机械臂 柔性关节 时滞 位置/力控制 神经网络控制 滑模控制 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]