垂直三连杆欠驱动机械臂通用控制策略设计

General control strategy design for vertical three-link underactuated manipulators

中文关键词  垂直三连杆欠驱动机械臂  轨迹  粒子群优化  跟踪控制  稳定控制

英文关键词  vertical three-link underactuated manipulator  trajectories  particle swarm optimization (PSO)  tracking control  stable control

中文摘要

      本文针对一类含单一欠驱动关节的垂直三连杆欠驱动机械臂提出一种基于振荡衰减轨迹的通用控制策略. 与传统的分区控制策略相比, 本文控制策略无需采用分区方式就能快速地实现将机械臂末端点由垂直向下初始位 置开始移动, 并最终稳定在垂直向上目标位置的控制目标. 首先, 根据驱动连杆的初始和目标状态, 为驱动连杆规划 含可调参数的振荡衰减轨迹. 该轨迹能够在一定调节时间内将驱动连杆直接由初始状态移动至目标状态. 基于连 杆状态间的耦合关系, 利用粒子群优化算法优化轨迹参数使欠驱动连杆在相同调节时间内也运动至目标状态. 接 着, 利用滑模方法设计跟踪控制器使驱动连杆跟踪优化后的振荡衰减轨迹, 这样, 系统末端点将由初始位置移动至 目标位置. 进一步利用极点配置方法设计镇定控制器克服重力的作用将末端点稳定在目标位置. 最后, 通过仿真实 验验证所提控制策略的有效性.

英文摘要

      This paper proposes a unified control strategy based on oscillatory attenuation trajectory for a three-link vertical underactuated manipulator with a single underactuated joint. Compared with the traditional partition control strategy, the proposed control strategy realizes the control target, which is to move the end point of the manipulator from the vertical downward initial position and to stabilize it at the vertical upward target position, quickly without using the partition control method. Firstly, according to the initial states and final states of the active links, the oscillation attenuation trajectories with adjustable parameters are planned, along which the active links can move from the initial states to the target states directly within a certain adjustment time. Based on the coupling action among the states of all links, we optimize the trajectory parameters by the particle swarm optimization algorithm so as to ensure the passive link reaches the target states in same adjustment time. Next, we design the tracking controller to track the optimized trajectories of the active links, moving the endpoint to the target position. Further, the pole assignment method is used to design the stabilization controller to overcome gravity and stabilize the endpoint at the target position. Finally, the effectiveness of the proposed method is verified by simulation experiments.