机械手轨迹跟踪精度控制仿真研究  

Simulation Study of Trajectory Tracking Precision Control for Robot Manipulators

作　　者：梁伟平[1] 王佳荣[1] 

机构地区：[1]华北电力大学自动化系,河北保定071003

出　　处：《计算机仿真》2015年第10期314-317,共4页Computer Simulation

摘　　要：由于机械手模型具有多自由度的关节机械结构,且关节之间强耦合,为提高跟踪精度,研究有效解除各关节之间耦合,实现机械手轨迹跟踪控制的稳定性和鲁棒性。通过将耦合严重的二自由度机械手进行解耦,使之变为2个二阶积分子系统。然后对上述子系统设计采用近邻L-M算法的神经网络α阶逆系统控制器。仿真结果表明:复合控制系统不但可实现机械手的线性化解耦控制,而且控制过程中仅用位置反馈就可以实现机械手的快速轨迹跟踪,在不需要改变控制器参数的前提下,能对不同频率、不同波形乃至不同幅值的曲线实现精确跟踪。Since most of the robot manipulators model have multiple degrees of freedom joint mechanical structure, and strong coupling between the joints, it is necessary to research a trajectory tracking control method which can effectively relieve the coupling between the Joints of robot. This paper makes the robot manipulator system into a two second - order integral subsystem by decoupling the two degrees of freedom manipulator systems. Then we design a controller based on nearest - neighbor L - M algorithm neural nets a - order inverse system. Simulation results show that, this complex control system can not only achieve linear decoupling control, but also achieve a rapid trajec- tory tracking only with position feedback. No need to change the controller parameters, the proposed controller can achieve accurate tracking of different frequencies, different waveforms and even different amplitude curves.

关 键 词：机械手 线性化解耦 神经网络逆系统 轨迹跟踪 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]