Please tell me about PID parameter setting
Thread Content
Dear masters, I just started to adjust the PID parameters. When setting the tower tank liquid level and the production flow rate in series, the situation as shown in the figure appeared. Could you please tell me which parameter in the PID is problematic at this time. Attached is the liquid level trend chart and the pid parameters of the main loop and the secondary loop (the flow pid parameter setting is OK and the automatic control is normal).
Can anyone help me?\ud83e\udd7a
The pid parameter, if different control objects have different experience values, can be adjusted purely proportionally until it oscillates with the same amplitude, record the proportion and period, then calculate the experience value according to the formula, and then fine-tune it according to the trend.* There is a problem in the question set. I have an experience to share. When the parameters are fine-tuned, the attenuation ratio does not necessarily follow the rules and will suddenly change. (This is all personal experience. The level is limited and is for reference only. If you don’t like it, don’t criticize it.)
This post was last edited by jlshnlhj on 2021-7-4 11:19. The picture of the red curve cannot see the time axis data, and there is no MV curve, and the K and T parameters (diameter of the tower kettle, regulating valve size) are not known. If you only follow the 5-minute time axis of the PV red curve in the above picture, it is estimated that the proportional effect is strong. You can appropriately reduce the proportional effect (2.5 to 2) and increase some integration time (1.2 to 2 minutes) to try.
The red curve is within 3 hours. I tried increasing the integral but the effect was not obvious.
This is how I debugged it, but there was no problem with the automatic control. However, this phenomenon occurred when it came to cascade control.
Try the differential adjustment parameters. The adjustment effect is not only affected by the PID parameters, but also by the matching of the controlled equipment and PID. This unstable control affects the material balance.
Will it be difficult to operate by adding differential control to the control of liquid level?
You have to look at the main and auxiliary curves. This is the main loop curve. What does the auxiliary loop look like? Is it a constant amplitude curve that is staggered from the main ring?
Why didn’t you mention reducing the proportional effect (changed from 2.5 to 2)? Try it. Don't say "increasing the integral has no obvious effect", what about the curve? Now you can only view the curve adjustment parameters. In addition, the concepts of integral action and integration time must be clarified. Increasing some integration time will reduce the integral effect.
Agree! Without the six parameters and curve time of the main and auxiliary circuits SV, PV, and MV, it is difficult to judge.