A2.3 Exasmoc Controller modes of operation-1

优化 – 理解主控制器的经济优化功能，以最具成本效益的方式运行流程，同时达到设定点或者最大限度地减少能耗/最大限度地提高产品收率等是非常重要的。基于个人成本的经济系数可以分配诸如能源成本或进料最大化等因素来实现最佳的经济效益。
在许多应用中，要求不能通过使用子控制器可用的所有自由度将所有变量保持在约束内控制。人们可以通过经济目标函数来使用额外的自由度，以便控制器除了控制之外经济地优化工艺。
经济函数是控制器最小化的双线性函数。它由操作输入和测量输出变量的双线性组合组成。经济系数出现在双线性组合中。
可以定义多个经济函数，但一次只能有一个经济函数活跃。可以在线选择活跃经济函数。经济系数也可以在线更改。
被控变量始终优先于经济函数。
为了实现经济优化过程，Exasmoc控制器允许用户定义经济函数。

Fig. A 2.5 Optimization, Prediction & Control function 图A 2.5优化，预测和控制函数

Fig. A 2.6 Deriving Economic outputs 图 A2.6派生经济输出

Fig. A 2.7 Example of optimisation structure 图A 2.7优化结构示例

控制目标：
 将C5的丁烷保持在设定点；
 保证塔盘在整流部分不液泛淹塔；
 将再沸器保持在最小负荷以上。
经济目标：
 最小化塔压（分离效率）；
 最大化FCU MCR排热量（增加进气量.....）
 尝试实现经济与调度部门设定的目标RVP。

A2.3 Exasmoc控制器操作模式

主控制器

 OFF

Exasmoc多变量，模型预测和优化函数为OFF。用户尝试使用基于DCS的BRC和ERC功能来稳定流程。

 STANDBY

Exasmoc处于待机模式（计算但不提供输出）。
使用基于DCS的BRC和ERC功能来稳定流程。用户正在观察Exasmoc standby操作，以便可以将以下模式投用。

 CONTROL

Exasmoc处于多变量模型预测控制模式下，并在设定点或指定设定范围内严格控制CV，同时在定义的约束内保持其他参数。基于DCS的BRC和ERC也在DCS PID等远程模式下接受来自Exasmoc的MV。

 OPTIMISE

Exasmoc处于多变量模型预测和优化控制模式下，在严格控制CV和其他控制目标的前提下，还追求经济有效的稳态运行模式。基于DCS的BRC和ERC在运行，且DCS PID等在远程模式下接受来自Exasmoc的MV值。可以为个人经济函数指定成本/收益部分，以实现优化的最大经济效益。

---

原文：
Optimisation – It is important to understand economic optimization function of the main controller in order to run the process at most cost effective way while achieving the set point or the desire to minimize/maximize energy consumption, product yield etc. Economic Coefficients based on individual cost factor such as cost of Energy or the Feed maximization can be assigned to achieve most optimum economic benefits.
In many applications, the control requirement to keep all variables inside the constraints does not use all the degrees of freedom available to sub-controllers. One can use the extra degrees of freedom by means of economic objective function so that the controller optimizes the process economically in addition to controlling it.
The economic function in is a bilinear function that the controller minimizes. It is made up of terms that are bilinear combinations of manipulated input and measured output variables. Economic coefficients appear in the bilinear combinations.
Multiple economic functions can be defined, but only a single economic function can be active at a time. The active economic function is selected on-line. The economic coefficients can also be changed on-line.
Controlled Variables have always a higher priority than Economic Function.
In order to achieve economic optimization of the process, Exasmoc controller allows user to define Economic Function.
Control Objectives:
 Keep C5 in Butane at setpoint
 Keep tray loading in rectifying section < flooding
 Keep reboiler above min load
Economic Objectives:
 Minimise column pressure (separation efficiency)
 Maximum heat removal from FCU MCR (increase crude intake .....)
 Try to achieve target RVP set by Economic & Scheduling dept.

A2.3 Exasmoc Controller modes of operation

Main Controller

 OFF

Exasmoc Multivariable, Model Predictive and Optimization function is OFF. User trying to stabilize process using DCS based BRC and ERC function.

 STANDBY

Exasmoc in a stand by mode (Computing but not providing outputs)
DCS based BRC and ERC function stabilizes the process. User is observing Exasmoc standby actions so that following mode can be put on line.

 CONTROL

Exasmoc in a Multivariable Model Predictive Control mode and tightly controlling CV at set point or within specified set range while maintaining other parameters within defined constraints. DCS based BRC and ERC in operation also DCS PID’s etc in a remote mode to accept MV’s from Exasmoc.

 OPTIMISE

Exasmoc in a Multivariable Model Predictive and Optimizing Control mode and tightly controlling CV’s and other control objectives in addition to cost effective steady state operation. DCS based BRC and ERC in operation, also DCS PID’s etc in a remote mode to accept values of MV’s from Exasmoc. It is possible to assign cost/benefits constituents for individual Economic Function to achieve Maximum Economic benefits from Optimization.

---

2017.3.26