基于一种加速PD型迭代学习控制算法的机械臂轨迹跟踪

针对二自由度机械臂非线性系统,迭代学习控制(iterative learning control,ILC)对于具有重复运动特性的机械臂有较好的控制效果。在扰动的情况下,设计了一种PD型迭代学习控制律,随着系统迭代次数的不断增加,通过在区间内对增益矩阵进行实时修改来缩短所需的修正区间,进而达到加快收敛速度的目的。首先,结合λ范数分析ILC的收敛性。其次,通过仿真验证所提出控制策略的可行性和有效性。最后,在相同条件下,仿真结果表明,PD型ILC收敛速度比P型ILC更快;带有扰动的PD型ILC比传统扰动型PD控制收敛效果更好。

A Novel Trajectory Tracking Approach forManipulator via an Accelerating PD-type

For the 2-DOF manipulator nonlinear system, iterative learning control (ILC) is very effective for the manipulator with repetitive motion properties. This paper proposes PD-type iterative learning control law with disturbance is designed. With the iterations increasing, the interval is changed by real-time correction of the learning gain matrix in the interval, in order to accelerate the convergence rate. First, the convergence of the ILC is analyzed in combination with the -norm and the corresponding convergence conditions. Second, the feasibility and effectiveness of the proposed are verified by numerical simulation. Finally, under the same conditions, the simulations show the convergence rate of PD-type ILC is faster than the P-type ILC; the PD-type ILC with disturbance has better convergence effect than the traditional PD control.