Hi Dr.Hedengren - Is this an example of additive feedforward control? Is so, how would you program in Python multiplicative feedforward control? Does your feedforward gain change between additive versus multiplicative feedforward control?
I know it has been a lot of time, but how come that for determining Kp value (delta u) valve opening percentage was chosen and not change in inlet flow? How to determine what is delta u value in general is it controlled variable or is it a consequence of controlled variable on real variable? And for everything you did with this videos and whole course, hats off to you, amazing work!
The actuator is often the change in a valve position (Delta u) or change in a pump speed. If there is an inner loop that controls flow then the primary controller has a control output of flow set point to the secondary controller. The secondary control has a controller output of valve actuator position (Delta u or valve lift). The process gain is the change in output (Delta y) divided by the change in input (Delta u). There is additional information here: apmonitor.com/pdc/index.php/Main/FirstOrderSystems
Hello Dr. Hedengren - as always thank you for posting your coursework. In the simulation you run the ODE solver with a time span of (0, 0.1). Is this intentional or a typo? Seems like it should be (0, 1.0).
Thanks for catching that! I increased A=5.0 m^2 (previously 0.5 m^2) for the cross sectional area and changed the time interval of integration to [0,1.0] so that the results stay the same. I fixed it in the step response code and the solutions that are posted at apmonitor.com/pdc/index.php/Main/FeedforwardCascadeControl
+Vegard S., that is the best practice to disable the outer loop. In this case, the outer loop didn't change the inner loop set point over the first 20 sec so I just kept it closed loop.
Hi Dr.Hedengren - Is this an example of additive feedforward control? Is so, how would you program in Python multiplicative feedforward control? Does your feedforward gain change between additive versus multiplicative feedforward control?
I know it has been a lot of time, but how come that for determining Kp value (delta u) valve opening percentage was chosen and not change in inlet flow? How to determine what is delta u value in general is it controlled variable or is it a consequence of controlled variable on real variable?
And for everything you did with this videos and whole course, hats off to you, amazing work!
The actuator is often the change in a valve position (Delta u) or change in a pump speed. If there is an inner loop that controls flow then the primary controller has a control output of flow set point to the secondary controller. The secondary control has a controller output of valve actuator position (Delta u or valve lift). The process gain is the change in output (Delta y) divided by the change in input (Delta u). There is additional information here: apmonitor.com/pdc/index.php/Main/FirstOrderSystems
Hello Dr. Hedengren - as always thank you for posting your coursework. In the simulation you run the ODE solver with a time span of (0, 0.1). Is this intentional or a typo? Seems like it should be (0, 1.0).
Thanks for catching that! I increased A=5.0 m^2 (previously 0.5 m^2) for the cross sectional area and changed the time interval of integration to [0,1.0] so that the results stay the same. I fixed it in the step response code and the solutions that are posted at apmonitor.com/pdc/index.php/Main/FeedforwardCascadeControl
What book are you using?
It is all online content available from apmonitor.com/pdc
Why dont you disable your outer loop when tuning the inner loop?
+Vegard S., that is the best practice to disable the outer loop. In this case, the outer loop didn't change the inner loop set point over the first 20 sec so I just kept it closed loop.