Control Views - Question of the week

Question #3 - What is "Standard" form of PID equation?
November 28, 2003

Question:

I have seen the PID equation expressed as:

wpe2D.jpg (3296 bytes)

I have also seen it expressed as:

wpe2E.jpg (3359 bytes)

Note the parenthesis. Which is correct? Is there a standard PID algorithm? Wouldn’t use of the second form (with parenthesis) cause the gain (Kp) to be interactive with the other terms?

Buy and download my eBook: The PID Control Algorithm: How It Works, How To Tune It, and How to Use It. 2nd ed. 62pp
Includes expanded information about cascade, feedforward, and other control schemes.
Adobe .pdf file                      only $15

visa-mclogo.gif (2156 bytes) Most major credit cards accepted.
Get more details, buy, and download
ControlSim  The PID controller and process simulator based on MS Excel spread sheets.    $45.00 Most major credit cards accepted.
Get more details, buy, and download

Answer:

There is no “standard” or correct PID algorithm. However, the second (Kp outside of parenthesis) form is the most common. Also, most standard tuning methods, such as Ziegler-Nichols and others, are based on that form.

As to the interaction, the parenthesis form of the equation actually reduces interaction. Using that form the reset (Ki) and derivative (Kd) parameters. The reason is that this form closely matches the process. Using this form, the Ki and Kd parameters relate only to dynamics of the process, and the Kp (proportional gain) relates to the gain of the process.

This is demonstrated by the Closed Loop Ziegler-Nichols tuning method. A controller is tuned with no or minimum reset rate (longest possible reset time) and no or minimum derivative. The gain is increased while the process is disturbed to find the gain at which the controller oscillates with constant oscillations-neither shrinking nor growing. That gain and the period of the oscillation is noted. The desired gain of the controller is calculated from the gain producing the constant oscillation. Ki and Kd are calculated from the period. In other words, if the dynamics of the process changes, using the standard algorithm (Kp affects all terms), Ki and Kd are changed. If the gain of the process changes, only Kp is changed.

 

Other Questions::

Difference between series and parallel derivative and integral
Derivative on process rather than error
Proportional vs. Proportional-reset for level control.
Reset windup and the problems it can cause.
Best scan time for loops.
Velocity and Full Value forms of the PID algorithm.
Interactive and non-interactive integral and derivative.
Fail safe level switch connections.
What is the "standard" PID equation? Gain affects all terms?
What are the benefits of realistic 3-D graphics for operator interface?
Should derivative be used on temperature loops?

Back to Control Views

Page Updated 03/26/04 09:19