基于干扰观测器的机械臂广义模型预测轨迹跟踪控制

为实现对多自由度机械臂关节运动精确轨迹跟踪，提出一种基于非线性干扰观测器的广义模型预测轨迹跟踪控制方法。针对机械臂轨迹跟踪运动学子系统，采用广义预测控制（Generalized Predictive Control,GPC）方法设计期望的虚拟关节角速度。对于机械臂轨迹跟踪动力学子系统，考虑机械臂的参数不确定性和未知外界扰动，利用GPC方法设计关节力矩控制输入，基于非线性干扰观测器方法实时估计和补偿系统模型中的不确定性。在李雅普诺夫稳定性理论框架下证明了机械臂关节角位置和角速度的跟踪误差最终收敛于零的小邻域。数值仿真验证了所提出控制方法的有效性和优越性。

Disturbance Observer-Based Generalized Model Predictive Trajectory Tracking Control for Robotic Manipulators

In order to realize the accurate trajectory tracking of the multi-degrees of freedom robotic manipulators in the joint motion space,a robust generalized predictive control method based on the nonlinear disturbance observer method is proposed.For the kinematic subsystem of the robotic manipulator trajectory tracking missions,the desired angular velocity trajectory is developed by the generalized model predictive control theory.Then,considering the parametric uncertainties and unknown external disturbances for the dynamic subsystem of the robotic manipulator trajectory tracking missions,the GPC method is adopted to design the torque control input,the nonlinear disturbance observer is employed to compensate the lumped unknown perturbations in the dynamics.Under the framework of Lyapunov stability theory,it is proved that the joint angle trajectory tracking errors and angular velocity tracking errors ultimately converge to the small neighborhoods of zero.The effectiveness and advantages of the proposed method are finally verified by numerical simulations.