六自由度机械臂约束预测控制系统的设计

Design of constrained predictive control system for 6 degrees-of-freedom manipulators

中文关键词  模型预测控制  机械臂  运动控制  前馈控制  动力学约束

英文关键词  model predictive control  manipulators  motion control  feedforward control  dynamic constraints

中文摘要

      本文提出了一种基于约束预测控制的机械臂实时运动控制方法. 该控制方法分为两层, 分别设计了约束预测控制器和跟踪控制器. 其中, 约束预测控制器在考虑系统物理约束的条件下, 在线为跟踪控制器生成参考轨迹; 跟踪控制器采用最优反馈控制律, 使机械臂沿参考轨迹运动. 为了简化控制器的设计和在线求解, 本文采用输入输出线性化的方式简化机械臂动力学模型. 同时, 为了克服扰动, 在约束预测控制器中引入前馈策略, 提出了带前馈$-$\linebreak 反馈控制结构的预测控制设计. 因此, 本文设计的控制器可以使机械臂在满足物理约束的条件下快速稳定地跟踪到目标位置. 通过在PUMA560机理模型上进行仿真实验, 验证了预测控制算法的可行性和有效性.

英文摘要

      Based on constrained predictive control, this paper proposes a real-time motion control scheme for manipulators. The control scheme is composed of two layers, with constrained predictive controller in the higher layer and trajectory tracking controller in the lower one respectively. The constrained predictive controller creates a trajectory with consideration of the system constraints for the lower layer controller. While the trajectory tracking controller adopts an optimal state feedback law to minimize the predicted tracking error. Input-output linearization is used to simplify the dynamic model used in the model predictive control (MPC) controller design in both layers. Furthermore, a feedforward-feedback control structure is proposed where feedforward is introduced to the constrained predictive controller to reject disturbance. In this way, fast and stable tracking as well as satisfaction of input constraints can be achieved. The proposed control scheme is tested for a industrial manipulator PUMA560 in MATLAB platform, and the simulation results show the feasibility and effectiveness of the proposed predictive control method.