机器人操作器系统自适应控制算法设计

ADAPTIVE CONTROL ALGORITHMS DESIGNS FOR ROBOT MANIPULATOR SYSTEMS

由于不可避免地存在负载、摩擦及间隙等因素的影响,常见机器人操作器系统的数学模型是一组系数未知的非线性微分方程。本文根据该数学模型,借助于对输入转矩的特殊补偿,首先导出了一种新的自适应控制算法,使常见机器人操作器系统的状态误差在幅度上比用常规自适应控制算法得到的误差要小;为了减小笛卡尔(Cartesian)坐标误差,还引入一种雅可比(Jacobian)关系,并将这种关系用于李雅普诺夫(Lyapunov)直接法设计,由此而设计的自适应控制算法进一步使系统的笛卡尔坐标误差减小。计算机仿真结果证明了本文所设计自适应控制算法的优良性能。

A set of nonliner differetial equations that represent a mathematical model of ageneral robot manipulator whose coefficients are unknwn due to the effect of payload,frictionand/or backlash,etc.,is considered.It is shown that by proper compensation of the inputtorque,the norm of the state error becomes less than that which resulted from theconventional adaptive control algorithems.A Jacobian relation is introduced and is used in thedesign methodology of the Lyapunov direct method to reduce further the cartesian error.Theoutstanding performances of control algorithms developed in this paper are demonstrated bycomputer simulation results.