Nonlinear Discrete-Time Control Approaches for Underactuated Manipulators
Nonlinear Discrete-Time Control Approaches for Underactuated Manipulators
复制标题
欠驱动机械臂的非线性离散时间控制方法
DOI：
10.1109/cdc.2005.1582851
复制
发表时间：
2005
期刊：
影响因子：
--
通讯作者：
H. Abdellatif
中科院分区：
文献类型：
--
作者：
Dirk Weidemann;Norbert Scherm;B. Heimann;H. Abdellatif
针对具有被动关节的水平平面欠驱动机械手，提出了两种非线性离散时间控制方法。第一种控制方法是基于扩展线性化的。它可以应用于不同类型的欠驱动系统，如小车-杆系统或高度欠驱动机械手。如果机械手没有制动机构，并且比主动关节更被动，则称为高度欠驱动机械手。然而，只有在误差足够小的情况下，上述控制方法才能很好地执行。因此，还提出了一种基于非线性在线优化的离散时间反馈控制器来抵消较大的轨迹偏差。为了减少计算时间，我们提出了一种考虑两个控制器之间的关系的非线性在线优化的初始化策略。文中给出了两种控制器的实验结果，验证了两种控制器的优缺点。
In this paper, we propose two nonlinear discretetime control approaches for underactuated manipulators operating in the horizontal plane which are equipped with passive joints. The first control approach is based on an extended linearization. It can be applied to different kinds of underactuated systems such as a cart-pole system or a highly underactuated manipulator. A manipulator is called a highly underactuated manipulator if it has no braking mechanisms and more passive than active joints. However, the control approach mentioned above performs well only if the errors are sufficiently small. Therefore, a discrete-time feedback controller based on nonlinear online optimization is proposed also to counteract large trajectory deviations. In order to reduce the computation time, we present an initialization strategy of the nonlinear online optimization considering the relation between both controllers. Experimental results of both controller are reported to demonstrate advantages and drawbacks.