高阶参数优化迭代学习控制算法

High-order parameter-optimization iterative learning control algorithm

中文关键词  迭代学习控制  参数优化  单调性  离散系统  线性系统  高阶

英文关键词  iterative learning control  parameter-optimization  monotonic convergence  discrete system  linear system  high-order

中文摘要

      针对线性时不变离散系统的跟踪问题提出一种高阶参数优化迭代学习控制算法. 该算法通过建立考虑了多次迭代误差影响的参数优化目标函数, 求解得出优化后 的时变学习增益参数. 从理论上证明了: 对于线性离散时不变系统, 该算法在被控对象不满足正定性的松弛条件下仍可保证跟踪误差单调收敛于零. 同时, 采用之前 多次迭代信息的高阶算法具有更好的收敛性和鲁棒性. 最后利用一个仿真实例验证了算法的有效性.

英文摘要

      A high-order parameter-optimization iterative learning control algorithm is presented for solving the tracking problems of a class of linear time-invariant discrete system. The proposed algorithm is based on a quadratic performance objective function with the tracking errors from earlier trials. By solving this function we obtain the optimal time-varying parameters as the learning gain of the iterative update law. It is proved theoretically that when applied to the relaxed linear discrete system, the proposed algorithm guarantees the tracking error to converge to zero monotonically even the original system is nonpositive. Moreover, since more information of previous iterations is considered in the proposed algorithm, the robustness and convergence performance of the algorithm are improved accordingly. Finally, a case study is carried out to illustrate the performance of this new algorithm.