Approximate FOPDT Dynamics with Graphical Fit

Dynamic processes are often characterized by a gain (Kp), time constant (tau), and dead time (theta). These exercises use a graphical fitting method to estimate the three characteristic parameters of the following dynamic systems described by a first-order linear system with time delay. See apmonitor.com/p...

[@lumek88]

Have you ever used a Sine Sweep for system identification?

[@apm]

That is another great way to generate data for system identification, especially for systems that have fast dynamics.

[@xursssulaxx]

What if you are given the time constant, and you’re asked to find the dead time?? But you don’t know t_0.632, however, you know what y equals at that time. But you don’t have a function for y(t)…you just have a transfer function written as “Kexp(-theta*s)/(tau*s+1)”? I’ve calculated K and tau, and all i know is there is a unit step of 10 and it takes 20 minutes to reach the new steady state. And the problem is asking for theta…

[@apm]

Here is additional help on graphical fit: apmonitor.com/pdc/index.php/Main/FirstOrderGraphical and optimized fit: apmonitor.com/pdc/index.php/Main/FirstOrderOptimization

[@vitalin.8581]

could you please say why the number is exactly 0.632? how do we get it?

[@apm]

Here is more information on why 63.2% apmonitor.com/pdc/index.php/Main/FirstOrderSystems

[@yannyepnang7645]

Please for problem what is true?
Teta p = 1,2 or teta p = 2,2?

[@apm]

For problem #1 it is 1.2 for the delay time. It starts from the step change.

[@philipasante8804]

You said the dead time is 1.2 secs and then when calculating for Tau p, you said dead time is 2.2.

[@apm]

Thanks for catching that.

[@TubOfLove69]

For problem 3, wouldn't the (Delta)Y*0.632=3.16, not 8 something??

[@apm]

Please see the comments starting at 11:52. It is (Delta)Y * 0.632 as I mentioned in the video. The answer that I show is correct.

[@jeyaneepan]

need Schrodinger equation solved by Python. Can you upload for us

[@apm]

This should help you: github.com/jakevdp/pySchrodinger There are many packages for Python and I haven't tested this one.

[@jeyaneepan]

thank you, but it does not contain vedio lecture

[@apm]

+Jeyaneepan Kanagaratnam, perhaps you could contact the author of the package if you can't get the details that you need from the documentation.