Modicon PID Loops

Application: PID Loop to control the flow in a large water line. Analog 4-20ma output feeding a large motor driven valve actuator. Input 4-20ma from a DP cell which is converted to flow.



Goal: Achieve a slow modulating output that will maintain a position. For example if a flow is achieved, then keep the valve at that %open position.



Issues: First of all, I have tried tuning the loop via PID parameters and the loop never seems to settle down. For example, if the Set point is 5 MGD, then the PV fluctuates quite a bit due to the valve constantly changing. It appears that the Analog Output is constantly changing with the PID. With this large motor driven actuator, I need the output to the valve to modulate slow and maintain position. In the past, I have tuned loops but they have been air actuated valves and it didn’t really matter how many times the valve actuated to achieve flow. This valve cannot actuate constantly due to the motor driven actuator and the size of the valve. It doesn’t appear that the adjustment of the PID parameters will allow me to achieve the slow modulating output and keep the valve in a maintained %open position. Note: The valve actuates extremely slow so there may be timing issues with the PID. I can set the loop in manual and adjust the output to maintain a steady flow rate.



There are items in the PID Loop that I’m not familiar with. What are the following items: Bias, High/Low Integral Windup, Initialization flag, Loop timer, error summation byte, Pointer to loop counter, max # loops scanned. Can any of these items help me achieve my goal? Do I have the PID setup correctly?



Is this not the right application for a PID?

The problem is bigger than you think

How do you know if the valve motor is stopped and not drifting if you don't have a position feed back?

With a valve position feed back you now have a PID to control position and a PID to control flow. See cascaded loops.

The outter PID compares the reference flow and the actual flow and generates a reference flow or position for the inner PID. The inner PID compares the valve position with the refence position and generates a control signal to the valve motor.

it will be somebody to help me with an example of a PID S7 300 or a ready program for turning in the simulator PLCSIM.
thank you.

rgive said:

Application:

There are items in the PID Loop that I’m not familiar with. What are the following items: Bias, High/Low Integral Windup, Initialization flag, Loop timer, error summation byte, Pointer to loop counter, max # loops scanned. Can any of these items help me achieve my goal? Do I have the PID setup correctly?
Is this not the right application for a PID?

Click to expand...

It should work OK perhaps your tuning constants are faster than the valve can respond so it's always chasing it's tail so to speak.
Turn the integral almost off and play with the gain, if it cycles just keep reducing it, could be as low as .2 or even lower once you get it settled start increasing the integral. No derivative should be required.
If you could sketch a curve and post it showing time along with your tuning parameters someone with more recent Modicon experience will be able to put you straight. A simple P&ID sketch of the process would help also.
High / Low integral windup are like limits on the control valve. If for example the source of water dried up the valve would wind open to 100%. If you know the maximum should be say 80% you set the High windup to 80% and thats where the valve will stop.
BTW another thing that will cause a valve to constantly cycle is slop in the linkage.
Good Luck
Roy

Roy, rgive posted that 6 years ago.

the other thing to think about is the length of pipework between the valve and the DP tranducer. This distance will cause some form of lag. This also will cause the PID controller some confusion as the control variable will not be having the desired affect on the process variable.

Mickey,
So he did, I'm a bit slow reponding I Guess
Klayton's post revived it.

Then again, I see my join date was April 2004, I could have sworn it was earlier this year.
Time flys when youre having fun
Cheers
Roy