冗余自由度仿人机械臂自适应迭代学习控制研究

Adaptive Iterative Learning Control of Humanoid Manipulator with Redundant Degree of Freedom

在深入研究自适应迭代学习控制理论、七自由度乒乓球机械臂动力学模型及轨迹规划的基础上,提出将改进后的自适应迭代学习控制算法运用到带有重复时变干扰的冗余自由度机械臂上。该控制系统旨在实现两大目标:一是使乒乓球机械臂准确快速地跟踪参考轨迹并在末点达到指定的击球速度;二是引入饱和函数减小输入转矩的抖振。Lyapunov理论分析及MATLAB仿真验证了整个控制系统的有效性:当迭代次数增加时,跟踪误差关于有限时间区间内一致收敛到零;加快迭代学习的收敛速度,并消除抖振。

An improved adaptive iterative learning algorithm is applied to 7-DOF manipulator with repetitive time-varying disturbance after analyzing adaptive iterative learning control（AILC） method, dynamic model and trajectory planning of redundant Ping-Pong manipulator. It is aimed to aehieve; （ i ） forcing the Ping-Pong manipulator to track time-varying reference signal fast and correctly, and with assigned speed of shot at the end point; （ ii ） reducing the chattering problem of input torque by using saturation function. The effectiveness of the proposed control system is demonstrated by Lyapunov analysis and simulation results： when the number of iteration increases, the tracking error will converge to zero uniformly with respect to the finite time interval, and chatter phenomenon disappears.