搭建仿真研究的第一步是定義要應用于流程的干擾順序。如下所示，此干擾將通過Inlet Flow rate（進料流量速率） DV注入：
| 步驟| 部分 | 變量| 屬性 | 新值| 舊值| 持續(xù)時間(dT)|
| ------------- |:-------------:| -----:|
| 20| DV | DV_1| Ramp| 5.0 | 0| 0|
| 40 | DV | DV_1 | Ramp| 10.0 | 5.0| 0|
| 60 | DV| DV_1| Ramp| 15.0 | 10.0| 0|
| 80 | DV| DV_1| Ramp| 0| 15.0| 0|
| 100| DV| DV_1| Ramp| -5.0 | 0| 50|
| 250| DV| DV_1| Ramp| 5.0 | -5.0| 100|
接下來，我們希望通過運用流程的干擾次序運行仿真為這一流程控制器建立基本的行為。在這種基本情景下所有的MVs都可用于控制（遠程），所有的CVs都有相同的優(yōu)先級（Priority=1）。下圖顯示了基線仿真的結果。
CVs的初始值都為50。在第5步時控制器狀態(tài)由Standby（掛起）切換到Control（控制）。這一控制狀態(tài)的改變使得SMOCPro操作MVs，從而將CVs移到各自的設定值： CV1（紅色）移至60，CV2（綠色）保持在50，CV3移至40。
現(xiàn)在讓我們來分析當我們把各個操作變量（MVs）脫開SMOCPro（通過將相應的MV控制狀態(tài)設置為“Local”）時將會發(fā)生什么。接下來的實驗中，在第5步時所研究的MV將從控制問題移出，在第400步MV將移回控制問題中。為了說明當斜坡CV相互矛盾時正確指定CV優(yōu)先次序的重要性，我們?yōu)槊總€實驗考慮兩種不同的設置。在每個仿真圖中，左邊部分將顯示所有的CVs具有相同優(yōu)先級（在本例中所有CV的Priority =1）時的情況，而圖的右側將顯示交錯CV優(yōu)先級的運行結果（CV1,CV2,CV3的Priority分別為1,2,3）。
思考當MV_1是“Local”的情況。當控制器切換到“Control”模式時， MVs將嘗試動作以達到控制目標。然而，能控制CV_1的唯一一個MV是脫開控制的MV_1。下圖顯示了仿真結果。圖中左右兩邊都表明SMOCPro不能控制CV_1到它的設定值。左側是所有的CVs都具有相同優(yōu)先級的情況，然而，SMOCPro識別出了可以修改CV_1的唯一手段于控制中是不可用的，因此將CV_1從控制問題中移除。如圖所示，CV_1的可達到目標是一條長鎖紅線，且保持放松。如圖右側所示，當CV的優(yōu)先級不同時其行為是完全相同的。在這種情況下其是不相關的，因為唯一能夠調節(jié)CV_1的MV_1是不可用的。
原文：
The first step in building the simulation study is to define the disturbance sequence to be applied to the process. This disturbance will be injected via the Inlet Flow rate DV as follows:
|Step |Section |Variable| Attribute |New Value| Old Value| Duration(dT)|
| ------------- |:-------------:| -----:|
|20 |DV| DV_1 |Ramp |5.0 |0| 0|
|40 |DV| DV_1| Ramp |10.0| 5.0| 0|
|60| DV| DV_1| Ramp |15.0| 10.0| 0|
|80| DV| DV_1| Ramp| 0| 15.0| 0|
|100 |DV| DV_1| Ramp| -5.0 |0 |50|
|250 |DV |DV_1| Ramp| 5.0| -5.0 |100|
Next, we wish to establish the baseline behavior for this process with our controller by running a simulation applying the disturbance sequence to the process. In this baseline scenario all the MVs are available for control (on remote) and all the CVs are at the same priority (Priority=1). The figure below shows the results of the baseline simulation.
The CVs initially start at the same value of 50. The controller status is switched from Standby to Control at step 5. This change in control status results in SMOCPro manipulating the MVs to move the CVs to their respective setpoints: CV1 (red) moves up to 60, CV2 (green) stays at 50 and CV3 moves down to 40.
Now let us analyze what happens when we take each of the manipulated variables (MVs) out of SMOCPro by setting the corresponding MV Control Status to “Local.” For the next experiments, the MV under consideration is removed from the control problem at step 5 and at step 400 the MV is brought back into the control problem. To illustrate the importance of correctly specifying the CV priorities for competing ramps we consider two different setups for each experiment. Each simulation plot will show the case of having all CVs at the same priority (in this case Priority =1 for all CVs) on the left side and the right side of the figure will show the results of having staggered CV priorities, Priority = 1,2,3 for CV1, CV2 and CV3, respectively.
Consider the case when MV_1 is on “Local.” When the controller is turned to “Control” the MVs move to try to achieve the control objectives. However, the only MV that can control CV_1 is MV_1 which has been taken out of the control problem. The figure below shows the simulation results. Both sides of the figure show that SMOCPro cannot control CV_1 to its setpoint. The case on the left has all CVs at the same priority, however, SMOCPro identifies that the only handle that can modify CV_1 is not available for control and removes CV_1 from the control problem. The reachable target for CV_1 is shown as a long-dashed red line that keeps being relaxed. The behavior is identical to the case where the CV priorities are staggered shown on the right side. This is irrelevant in this scenario since the only MV_1 capable of regulating CV_1 is not available.
2016.5.27
