How to implement a PID Loops control system in PLC programming
I think,this article will be you like because many PLC developper want to understand in PID controller,acquisition and transmission for PLC .
As I have read and understood what is the proper way to tune a loop that the data on the response of a process of maintaining and tuning parameters (eg data: building a FOPDT) model to collect.
PID on PLC Lesson Resources
What is P in PID?
What is I in PID?
What is D in PID?
Proportional-Only Control - Why Won’t It Work?
Scaling with y=mx+b (chapter 1)
Scaling with y=mx+b (chapter 2)
Q: But in almost all the experiences I have had with tuning loops in the industry, I have never seen an engineer.
Either you have never used a real control engineer and the system is so simple that it can be manually adjusted. A real control engineer would know that more steps manually or comparison.
Q: So make me some questions, such things should realistically possible, hopefully this is not just my inexperience in the real world monkeys stuff of science.
As you said, the real problem is the data and installation instructions, a model. If the model is a FOPDT or SOPDT or neither. Position control models also come with an integrator that speed is not to integrate. I’ve always, you hope that the system free linear.
I’ve never used a system with a transfer function. For me all the systems I’ve seen are not designed and is kludged. This finding means that the control engineer has a transfer function empirically close enough.
Q: A mechanism should be terminated when the melody?
Yes, change systems.
Q: If not, when it is appropriate, whether there are still cases where it is counterproductive to do that?
If it takes long to do. I do not even enough time for a car tune-motion system to do. I can take hours to one power or another.
Q: Is there ever a point of diminishing returns in the efforts to try to make a loop like this song is based?
Yes, there is a point of diminishing returns in almost all things. We often see people spend too much time trying to perfect as it will never reach to reach. If the system meets your tolerances, so it’s good enough.
Q: Modeling of the system is good because it increases your understanding of how the system works and provides evidence for the next guy who comes. With time can be added to the model taking into account other things. Models better with time.
This is the biggest thing I remember from modeling. If you do not understand what you are trying to determine how effectively they can control?
The main advantage of this site is that if you model the calculation of the earnings test and their answer is quite simple.
For example (a first-order linear model with a left-plane poles very Temperature Control) applications entirely stable.
No matter how high your profits increase, the model does not fluctuate. But as you know, it is quite easy in the real world reason for what looks like a first order system fluctuate.
Why? This is where modeling can help. If you start your system, you can transport delay (dead time), then available to analyze a system FOPDT.
Update your model with this information. You’ll notice that your disk has a first order (or higher) and the reaction of a stick (or two) added.
You’ll notice that your sensor has a first order (or higher) and the response that half pole (or two) added.
All this information can be added, so your model more representative of your real system.
