Designing a Controller-Sub-controller Compaction Points Tab-2

压实点网格选项字段和按钮
Settling Time estimate (from model)（从模型中设定估计时间） 指示达到稳态所需要的时间（子控制器周期的倍数，分钟或秒）。
Calculator（计算器） 根据用户指定的输入时域和设定时间（见下文）重新计算压实点。
输入
ControlPeriod ×dT（控制周期×dT） 子控制器周期中的输入压实点列表。
Minutes/Seconds（分/秒） 表示成时间单位的输入压实点列表。
输出
ControlPeriod×dT（控制周期×dT） 子控制器周期中的输出压实点列表。
Minutes/Seconds（分/秒） 表示成时间单位的输出压实点列表。
Reset to Default（重置为默认） 将压实点重置为从操作输入与被控输出响应的动态模型中自动计算的结果。
注：稳态时间被限制为1000个子控制器周期以内。对较慢的子控制器动态，请增加子控制器周期。
用户可以通过在压实点表格相应栏中手动输入所需值来分配输入和输出压实点。这些值可以根据时间单位（mins/secs）或子控制器周期（Control Period × dT)的格式输入和/或查看。从mins/secs到子控制器周期单位间转换是自动的（反之亦然）。计算出的值将自动进行四舍五入，以圆整为子控制器周期的整数倍。
另一个可供用户使用的工具是压实点计算器，可通过单击Calculator按钮进入。这个工具可根据用户指定的输入时域和设定时间来确定合适的压实点。用户将打开下述输入框：

这两个可修改的字段分别是：
• Input Horizon（输入时域）- 该值作为一个（未来）时间窗口将用于生成预测动作规划。此外，该数也限定了最后输入压实点的位置。
• Settling Time（稳态时间） - 该值表示达到稳态所需的时间。该值与预估值（从模型中）相比较越大，动态控制过程将把更大重心放在稳态处理上。
注：输入时域是向动作计划提供自由度的一个有用手段。这个量对“浪潮控制”非常重要。
静态约束
在子控制器层，被控变量被定义为子控制器的MV和/或POVs的线性组合。在它的观点中，定义整个子控制器的控制目标（CV）是不允许的。子控制器使用它们的子模型（与其它子控制器的前馈一起）来优化动态目标。跨越子控制器的控制目标被处理成静态目标，并执行为稳态下的动态子控制器约束。

---

Compaction Points Grid Tab Fields and Buttons
Settling Time estimate (from model) An indication of the time needed to reach steady state (in number of sub-control periods, and minutes or seconds).
Calculator Recalculates the compaction points on the basis of user-defined input horizon and settling time (see below).
Input
*ControlPerioddT **The list of input compaction points in the sub-controller period.
Minutes/Seconds A list of input compaction points in time units.
**Output **
*ControlPerioddT **The list of output compaction points in the sub-controller period.
**Minutes/Seconds **A list of output compaction points in time units.
Reset to Default Reset the compaction points to those computed automatically computed from the response of dynamic model from the manipulated inputs to the controlled outputs.
NOTE: The time for steady state is limited to 1000 subcontrol periods. For slower sub-controller dynamics, increase the subcontrol period.
The user may assign input and output compaction points manually by entering the desired values in the corresponding cells in the compaction points grid. These values may be entered and/or viewed in either time-units (minutes/seconds) or in sub-controller periods (Control Period × dT). The conversion between the fields from minutes/seconds to sub-controller periods (or vice-versa) is automatic. The calculated numbers will automatically be rounded up so that the conversion gives an exact multiple of the sub-controller period.
Another tool available to the user is the Compaction Point Calculator accessed by clicking the Calculator button. This tool determines suitable compaction points based on user specified entries for the input horizon and settling time. The following input box opens:
The two modifiable fields are:
• Input Horizon – This value is the (future) time window over which a forecast of the plan of action is generated. In addition, this quantity also defines the location of the last input compaction point.
• Settling Time – This value denotes the time required to reach steady-state conditions. The larger this value is compared to its estimate (from model), the more emphasis is put on the steady state during dynamic control.
NOTE: The input horizon is a useful handle for providing freedom to the plan of action. This quantity is important for “surge volume control.”
Static Constraints
At the sub-controller level, the controlled variables are defined as linear combinations of that sub-controller’s MVs and/or POVs. At his point, defining a control objective (CV) across sub-controllers is not allowed. Sub-controllers optimize their dynamic objectives using their sub-model (together with feed forward from other sub-controllers). Control objectives across sub-controllers are processed as static objectives and are enforced as constraints on the dynamic sub-controllers at steady state.

---

2016.9.29