The block diagram is simple.
The Control is the big problem: Since you have different water levels, the Tr. Function-and Parameters are different.
You are in the case of MIMO structure control, but the worst case - cross parameters (+ signals -level and temperature) disturbances.
(Theory recomendes to build feed back to compensate the cross links, and to isolate the channels Input/Out for Level and temperature).
Let's presume that we have 10KW warming Input for the system.
1. For a Level representing one liter water, the time constant of the Object will be 10 seconds.
2. For a lever of 5 litters water the time constant will be already 500 seconds...
and so on... reaching the moment to decrlare the object with huge pasrameters variations (proportional??? NO!) to the level-quantity-energy accumulated-inertia+++.
The control design must be done for the fastes object transfer function, i.e. 10 seconds. For the rest - complicated control design requesting measuring the level and correcting the P and Integral part of the PID.
Stability must be covered for the fastes case. the rest is secure with the approxiamtion of 5%.
No fast changing of the parameters are allowed - singular disturmancy leading to unstability, and BOOOM! (I have seen a dead engineers !)
But the problem is with the tepperature dynamic: Normaly the dynamic of the temperature will drastically affect the post connected processes. This is/are the process after the Water-Level/Temp.
First you have to start with the identification of the object:
1. leveling the water into 100 levels and extracting the dynamic responce, as
2. Step Heating - measuring the Temperature time dynamics, and taking the Time Constante.
3. After that you have to return to the precision of the Level control (Only a poportional feed back will make it - the object -level is small I order lag + huge Integral block description).
And to take it from there...
Give me a call...
Facebook
Google
GitHub
Linkedin
