SMOCPro中的阀门饱和处理—dSPdV-4

2.将dSPdVLo/Hi增益的量纲设置为等于MV范围：

这样做最重要的是要保证在界限时增益具有正确的符号。
4.监测并比较计算的MV限和用户指定的SPLo和SPHi限。如果这些值是不同的，必须特别注意采取措施处理差异。
注意事项：以这种方式设置的限制保证了阀位永远不会在DCS阀位高低限之外。然而，SMOCPro拥有的防止饱和和控制器饱和度的任何功能都将被禁用。
•最后，当我们控制一个OP且防止控制器饱和被视为关键时， OP可以明确地作为CV被添加到SMOCPro控制器的过程模型中。这一步可以通过使用图形化模型构建器（GMB）内的“变换”功能来表示非线性阀特性而大大受益。
SMOCPro dSPdV计算示例
我们现在通过例子来说明dSPdV计算。第一个例子突出了 “正常”操作状态的场景。第二个例子显示了当用了饱和PV时dSPdV计算的行为。最后，第三个例子显示了相对于dSPdV计算，为MV指定适当Max Move Size值的重要性。最终，我们提出了在处理OP所需的抗物理限时，两种仿真在计算SP限时行为的不同。
例1：正常操作
考虑图1（上文）描绘的具有以下特征的系统：

|低限 |高限|
| ------------- |:-------------:| :-----:|
|OPLo = 10| OPHi = 90|
|SPLo = 500 |SPHi = 1000|
|dSPdVLo = 20| dSPdVHi = 5|
当前的运行情况是在OP=60时，SP（回读）= PV = 800。
我们首先需要计算的是OP对SP的增益。根据用户指定的参数，在当前OP时增益为：

接下来我们通过计算等式（1a-b）获得内部限制：

最后，求解方程（2a-b）获得所计算的限制：

在这个例子中我们注意到，当OP = 60时控制器依然可以访问整个MV范围。传递到内核的计算值等于用户指定的这些值。
例2：饱和PV
现在我们来考虑当OP值在低于SPHi时饱和的情况。因此我们需要将控制器的SPHi带回PV以使其能做出任何相关动作。该系统的条件是：

|低限 |高限|
| ------------- |:-------------:| :-----:|
|OPLo = 10 |OPHi = 90|
|SPLo = 500| SPHi = 1000|
|dSPdVLo = 10 |dSPdVHi = 15|

---

原文：

2. Set the magnitude of the dSPdVLo/Hi gains equal to the MV range:
   In doing this, it is critical that the gains at both limits have the correct sign.
3. Monitor and compare the calculated MV limits vs the user-specified SPLo and SPHi limits. Should these values be different, special attention must be taken to address the discrepancy.
   Word of Caution: Setting the limits in this manner ensures that the valve position is never outside the DCS valve low and valve high limits. However, any capabilities that SMOCPro possesses to prevent wind-up and controller saturation will be disabled.
   • Lastly, when controlling an OP and preventing controller wind-up is deemed critical, the OP can be explicitly added to the process model in the SMOCPro controller as a CV. This step can greatly benefit by using the “Transforms” feature within the Graphical Model Builder (GMB) to represent nonlinear valve characteristics.
   **SMOCPro dSPdV Sample Calculations **
   We now present examples illustrating the dSPdV calculations. The first example highlights a “normal” operating regime scenario. The second example shows what the dSPdV calculation does when presented with a saturated PV. Lastly, the third example shows the importance of specifying an appropriate MV Max Move Size value with respect to the dSPdV calculation. Lastly, we present two simulations highlighting difference in behavior for the calculated SP limits when dealing with desired versus physical limitations on the OP.
   **Example 1: Normal operation. **
   Consider the system depicted in Figure 1 (above) with the following characteristics:

| Low Limits | High Limits|
| ------------- |:-------------:| :-----:|
| OPLo = 10| OPHi = 90|
| SPLo = 500| SPHi = 1000|
| dSPdVLo = 20 | dSPdVHi = 5|
The current operating conditions are SP(readback) = PV = 800 at OP = 60.
The first thing we need to calculate is the OP to SP gain. With the user-specified parameters, the gain at the current OP is:
Lastly, evaluate equations (2a-b) to get the calculated limits:
Next, we calculate the internal limits using equations (1a-b) to obtain:
In this example we notice that at OP = 60 the controller still has access to the full MV range. The calculated values passed to the kernel are equal to those specified by the user.
Example 2. Saturated PV.
Now consider the case when the OP is saturated at a value lower than SPHi. Thus we need to bring the SPHi to the PV for the controller to be able to do any relevant controller moves. The system conditions are:

| Low Limits| High Limits|
| ------------- |:-------------:| :-----:|
| OPLo = 10| OPHi = 90|
| SPLo = 500| SPHi = 1000|
| dSPdVLo = 10 | dSPdVHi = 15|

---

2016.7.5