P+I+D - http://www.hvacmechanic.com/ Forums


Original message

jbmendoza

210.14.44.170

"P+I+D" , posted Thu 11 Aug 23:09user profileedit/delete messagepost reply


can anyone explain in laymans term what is prop, P+I and PID control thanks

 


Replies:

superdave

65.30.111.206

"Re(1):P+I+D" , posted Sat 13 Aug 12:40user profileedit/delete messagepost reply


PID loop is a control algorithm,
P= Proportional(how much)
I= Integral(how long)
D= Derivative(sum of the two).

OK clear as mud?

Here is what it means in laymans terms, the PID controller sends out some sort of signal to adjust something, let's say a three way mixing valve (thats the P), the PID controller then waits for some amount of time to see what happened (thats the I).

So if the the three way valve is controlling a cold deck of a chilled water system to 55f supply air the PID maybe programmed to begin opening the valve if the supply reaches 57f, stop opening the valve if the supply drops to 55f and start to close the valve if the supply drops to 53f. Most controllers don't need the D (Derivative) to operate. Carriers E series RTU have the PID loop adjustable to allow the field to speed up or slow down the economizer open/close and even compressor stage up/down and loading on/off. I hope this helps..

Superdave

SuperDave

 

 

jocko

65.40.175.217

"Re(2):P+I+D" , posted Sat 13 Aug 13:01user profileedit/delete messagepost reply


This is one thing I actually know well, just not in HVAC terms!

PID describes the gains in a closed-loop servocontrol system. As the man said, proportional, integral, and differential. Any closed-loop or servo control is an error-driven system. That is, function is based upon the difference between the command and feedback signals. ERR = CMD-FDBK. In order to be responsive, the error signal is amplified by a factor known as the loop gain. These gains are P, I, and D based upon their mathematical properties.

Proportional gain is just that, an amplification directly proportional to the amount of error present at that time. In math, that is the specific point on the data curve.

Integral gain is an amplification based upon how long an error signal has been present. Mathematically, that is analogous to the area under the data curve. The intergral.

Differential gain is an amplification based upon the rate-of-change of the error signal. That is, the slope of the data curve. The differential.

In control systems, proportional gain by itself can be sluggish or perform poorly. An integral gain is used to eliminate a persistent "following error" or to hit final position. The integral will "wind up" over time. Differential gain is used to prevent overshoot or dampen out an unstable system. Getting the loop properly tuned is a bit of a crap shoot.

jocko