For industrial batch processes subject to time-varying uncertainties and nonrepetitive disturbances, this article proposes a robust static output feedback (SOF) based iterative learning control (ILC) design with a finite-frequency-range two-dimensional (2D) performance specification. An important advantage of the proposed design lies in its simple structure and easy implementation, with no need to store the state information of the learning controller as used in the existing 2D dynamic output feedback based ILC methods. Based on a 2D Roesser system setting, sufficient conditions are established in terms of nonlinear matrix inequalities to guarantee robust stability of the resulting closed-loop ILC systems along both the time and batch directions, by using the matrix dilatation technique. Moreover, a finite-frequency-range 2D performance specification for attenuating nonrepetitive disturbances is introduced for control synthesis. To derive the feasible SOF based ILC controller gains, a two-stage heuristic approach is developed for iterative computation based on predesigning the corresponding state feedback based ILC gains. An illustrative example is used to demonstrate the effectiveness and merits of the proposed method.

中文翻译：

---

基于鲁棒静态输出反馈的迭代学习控制设计，具有有限频率范围二维规范，用于受非重复干扰的批处理

对于受时变不确定性和非重复干扰影响的工业批处理过程，本文提出了一种基于稳健静态输出反馈 (SOF) 的迭代学习控制 (ILC) 设计，具有有限频率范围的二维 (2D) 性能规范。所提出设计的一个重要优点在于其结构简单且易于实现，无需像现有的基于 2D 动态输出反馈的 ILC 方法中那样存储学习控制器的状态信息。基于二维 Roesser 系统设置，通过使用矩阵膨胀技术，在非线性矩阵不等式方面建立了充分条件，以保证所得闭环 ILC 系统在时间和批次方向上的稳健稳定性。此外，有限频率范围的 2D为控制合成引入了衰减非重复干扰的性能规范。为了获得可行的基于 SOF 的 ILC 控制器增益，基于预先设计相应的基于状态反馈的 ILC 增益，开发了一种用于迭代计算的两阶段启发式方法。一个说明性的例子被用来证明所提出方法的有效性和优点。