基于奇异摄动理论的输入有界机器人轨迹跟踪控制

针对机器人轨迹跟踪中驱动扭矩有界的问题.提出一种基于奇异摄动系统稳定性理论的推广算法.通过在控制律中引入含有误差增益的饱和函数,保证扭矩输入绝对值的上界在给定限制范围内.并可通过适当调节误差增益系数改善系统的轨迹跟踪性能.同时,算法中采用仅包含位置跟踪误差信息的线性滤波函数产生用来替代真实速度误差的伪速度误差信号.使得整个系统的闭环控制不需测量转速.根据提出的推广算法,设计了一种全新的输入有界控制律,验证了算法的有效性.仿真试验对比结果表明,该算法能够严格保证扭矩控制输入的有界性,并在相同参数条件下相比于其他算法,具有更优的轨迹跟踪效果.

Trajectory tracking control for robot manipulators with bounded inputs based on the singular perturbation theory

To meet the requirement of a bounded driving torque in the robot trajectory tracking, we propose a generalized algorithm based on the stability theory of singularly perturbed systems. A saturation function with error gain is introduced in the control law, which guarantees the absolute value of the input torque to be upper-bounded within any specified range. Besides, a better tracking performance can be achieved by properly adjusting the error gains. In the meantime, a pseudo velocity error signal for substituting the actual one is generated from a linear filter containing only the position tracking error, thus, making the whole closed-loop control free from the velocity measurement. Based on the proposed algorithm, we design a novel control law with a bounded input. Comparison of simulation results show that the proposed algorithm can strictly ensure the specified bound for the inputs to the torque controller, and gives a better tracking result than the other ones under the same parameter conditions.