SEARCH Archives
Loading
Sponsored by:

Ask Control Engineering

Ask Control EngineeringThe Ask Control Engineering blog covers all aspects of automation, including motors, drives, sensors, motion control, machine control and embedded systems. Control Engineering answers questions from readers of Control Engineering's print and online magazines, newsletters and other publications. To comment on any blog posting, click on the post's highlighted question and scroll to the "Post a Comment" box at the bottom. Submit questions as comments to any existing post.

See all Ask Control Engineering blogs and comments


What's the problem with overshoot?

September 10, 2010

Bookmark and Share

Dear Control Engineering: I was looking at Back to Basics in the August issue about closed loop control. What is the main concern about overshoot? Why is it a problem, and can’t a controller prevent it?

An ideal controller should be able to move a process variable quickly to the desired setpoint without shooting past. Sometimes controllers will push the variable aggressively toward the setpoint knowing that it may go past, or overshoot the desired value and have to turn around and come back. In many applications, this isn’t a problem. For example, let’s say you’re trying to control flow. The sensor is a flowmeter and the controller is a valve. Assuming everything is sized appropriately, the valve should be able to accommodate the desired flow easily, so it’s no big deal if it goes a little above or below the setpoint because it can easily reverse its effort.

But, let’s say you’re trying to control the temperature of a large tank of liquid by modulating the amount of natural gas to a burner. If the liquid is too cold and you’re trying to bring the temperature up quickly, the controller may turn the gas all the way up. However, if you overshoot the setpoint, the controller can do nothing to make the liquid cooler. There is no reverse effort available. Eventually some of the heat will dissipate and some might be carried away by a cooler feedstock coming into the tank, but there may be no counter control effort beyond turning the burner off.

To complicate things further, let’s say the tank is filled with vanilla pudding and overheating will cause it to burn. You want it to boil, but not get too hot. In those situations, the controller has to be programmed in a way that it will not pass the desired point or the product could be ruined.

Controllers can be programmed to avoid overshoot, but this is usually at the expense of aggressiveness. It has to slow the control effort more and more as it approaches the setpoint, which drags out the time required to hit the desired value.

For a more thorough discussion of the topic, read the article Understanding Derivative in Process Control, especially the sidebar written by Vance VanDoren.

--Pete Welander, pwelander(at)cfemedia.com

Displaying results 1 to 1 out of 1
 

Erick

Friday, 17-09-10 12:21

Hello - I am working in the process to automate a pasteurizer. I am looking for a sensor technology to detect the change from water to orange juice and from orange juice to water in the output process.  It has a color sensor... but it doesn't detect pretty well the change. Do you know a better sensor? Do you know if today's pausteurizers come with one? Thank you.
Erick

 
 

CAPTCHA image for SPAM prevention


Poll of the Week

How integrated are your maintenance, controls, and instrumentation systems?
Fully integrated
Somewhat integrated
Loosely integrated
They're not


Click Here for Poll Archives
Sponsored by:

CFEMedia.com | Subscribe to Magazine | Advertise | Contact Us | About Us | Site Map | Privacy Policy
Channels | New Products | Media Library | Connect | Industry News | Events and Awards | Newsletters | Blogs | Magazine
Control Engineering | Plant Engineering | Consulting-Specifying Engineer
All content copyright © 2010-2013 CFE Media. All rights reserved.