In this article, a robust networked iterative learning control (ILC) method is presented for switched nonlinear discrete-time systems (NDTS) subject to non-repetitive uncertainties and random data dropouts. In the proposed robust networked ILC scheme, the switching law, iterative initial state, and disturbances, all of which vary with iterations, are well addressed. Corresponding to the actuator side and the measurement side of the networked switched NDTS, the random data dropouts occurred are compensated by the input signals at last iteration and the reference outputs, respectively. As a result, it is theoretically proved that under the non-repetitive uncertainties of the switched NDTS, the mathematical expectation of ILC tracking error remains bounded during the ILC process. While the non-repetitive uncertainties are progressively convergent in iteration domain, a precise tracking to the reference trajectory in mathematical expectation sense can be achieved. The effectiveness of the proposed networked ILC design is validated by a numerical example.

在本文中，针对受非重复不确定性和随机数据丢失影响的切换非线性离散时间系统 (NDTS)，提出了一种稳健的网络化迭代学习控制 (ILC) 方法。在提出的鲁棒网络 ILC 方案中，开关定律、迭代初始状态和扰动，所有这些都随着迭代而变化，得到了很好的解决。对应于网络化切换NDTS的执行器侧和测量侧，发生的随机数据丢失分别通过上次迭代的输入信号和参考输出进行补偿。结果，理论上证明了在切换NDTS的非重复不确定性下，ILC跟踪误差的数学期望在ILC过程中保持有界。虽然非重复不确定性在迭代域中逐渐收敛，但可以实现对数学期望意义上的参考轨迹的精确跟踪。所提出的网络化 ILC 设计的有效性通过一个数值例子得到验证。