In industrial processes, zone control is often applied to some control systems that do not have set-point control requirements. Zone control involves constraining controlled variables within a range of feasible solutions, so there is a tension between interval constraint intensity and interval feasibility. To meet the requirements of interval feasibility, general control schemes often soften constraint intensity, resulting in frequent constraint violations. However, for some irreversible processes such as chemical and biochemical processes, violating constraint can lead to unpredictable results. This study attempts to solve this problem through improving the performance index of general model predictive controls with soft constraints. This goal is achieved through introducing additional margin to controlled variables in order to strengthen control intensity without increasing computational complexity. This approach effectively reduced the frequency of zone violations and the size of output errors, and guaranteed zone feasibility. Furthermore, this approach was implemented without significant increase in energy consumption and actuator operation. The stability of the algorithm was proven using Lyapunov theorem. Comparative simulation results demonstrated the effectiveness of the proposed method compared with conventional methods.

中文翻译：

---

具有软约束附加余量的模型预测区域控制

在工业过程中，区域控制通常应用于一些没有设定点控制要求的控制系统。区域控制涉及在可行解的范围内约束受控变量，因此区间约束强度和区间可行性之间存在张力。为了满足区间可行性的要求，一般控制方案往往会软化约束强度，导致频繁违反约束。然而，对于一些不可逆的过程，如化学和生化过程，违反约束会导致不可预测的结果。本研究试图通过提高具有软约束的通用模型预测控制的性能指标来解决这一问题。这个目标是通过为受控变量引入额外的裕度来实现的，以便在不增加计算复杂性的情况下加强控制强度。这种方法有效地减少了区域违规的频率和输出错误的大小，并保证了区域的可行性。此外，该方法的实施并未显着增加能耗和执行器操作。利用李雅普诺夫定理证明了算法的稳定性。对比仿真结果证明了该方法与传统方法相比的有效性。这种方法的实施并没有显着增加能源消耗和执行器操作。利用李雅普诺夫定理证明了算法的稳定性。对比仿真结果证明了该方法与传统方法相比的有效性。这种方法的实施并没有显着增加能耗和执行器操作。利用李雅普诺夫定理证明了算法的稳定性。对比仿真结果证明了该方法与传统方法相比的有效性。