初态学习下广义系统的迭代学习控制算法

摘要

For a class of singular systems operated repeatedly on finite interval,the paper proposes an iterative learning control algorithm based on initial state learning with index gain for control input.By using the operator theory,the paper gives the convergence conditions.The algorithm can speed up the convergence rate on the basis of ensuring convergence when compared with the closed-loop D type algorithm.The effectiveness of the iterative learning control algorithm is proved by the illustrative example.%针对一类在有限时间区间上运行的广义系统,提出了一种初态学习下的迭代学习控制算法,该算法对系统的控制输入采用闭环带指数变增益D型学习律.基于算子理论,给出了算法的收敛条件.该算法与闭环D型算法相比,在保证收敛的基础上加快了算法的收敛速度.数值仿真结果说明了该算法的优越性.