不依赖模型的机器人鲁棒自适应跟踪控制(英)

Model-Independent Robust Adaptive Tracking Control of Robot Manipulators

提出了一种新颖的鲁棒自适应控制策略，用于不确定性机器人的轨迹跟踪．它不需要任何模型知识，唯一需了解的是系统的阶数和输出的位置及速度状态。理论和仿真均证明，系统的不确定性诸如摩擦力、外部扰动及未建模动力学带来的不确定性影响，均可被设计的控制律补偿，最后可保证全局指数收敛或全局一致最后有界的结果．另外，本文还给出了跟踪误差的暂态测量．

This paper proposes a novel model-independent robust adaptive control strategy fortrajectory tracking of robot manipulators with uncertainties,in which the only information requiredin establishing the strategy is the degree-of-freedom(DOF) and output state of the system. It isshown by theories and simulations that uncertain effects such as frictions and external disturbancesor unmodelled dynamices can be eliminatated and global exponential stability(GES) or global uniform ultimate boundedness (GUUB) stability can be guaranteed. Furthermore,we also give a measure of the transient tracking error performace.