When Should You Use Cascade Control?

Cascade control has many potential benefits:
1. It will catch certain load changes sooner and correct for them faster.
2. It will effectively linearize the secondary variable. 3. It can and should shorten the natural period of the primary loop. 4. Some instrumentation tricks can be played.
For more information on these subjects and others you may refer to a booklet "Controller Tuning and Control Loop Performance, a Primer" , which is available in the US for $17, postpaid if prepaid. It has sold 29,996 copies to date. Control Engineering Magazine has referred to it as "PID Without the Math".
Information may be found at http://members.aol.com/pidcontrol /. A page on Companion Software ($11) is there also.
If you wish, I could e-mail you the same information.
Straight-Line Control Co.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
St Clair's coverage of Cascade Control in his book is well worth a read; and so are his other articles. They are based on his solid experience on process plants. I bought the book and freely acknowledge the influence it has had on my work.
Also go to:
http://www.eurotherm.com/training/tutorial/tutor.htm
Here you will see a list of 34 (and growing) articles on temperature and power control for use by the man in the plant with other jobs and little time for mathematical analysis.
Article #4 there is on Cascade Control.
I am wondering why the system spills out so many repetitions of St Clair's posts?
Art
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Arthur Holland wrote:

For a lead on another great book read about my favorite at http://users.erols.com/jyavins/servo.html
Jerry
--
Engineering is the art of making what you want from things you can get.

  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I've found that cascade provides a slower, more stable reponse compared to direct control.
An example, an internal reactor temperature may be controlled by a steam heat exchanger/+10C Glycol circulated jacket. The internal temperature drives the setpoint of the circulated jacket.
This works well when you have both exothermic and slightly endothermic reactions that need to be controlled at a particular temperature.
I've also seen cascade work well in level control situations - the percentage level driving the setpoint of the exit flow controller (where the incoming flow is not a control point).
Where I have seen trouble is whenever poor engineering leads to competing cascade loops - one variable being affected by different control loops.
I am not an engineer - more of an operator turned inst. tech. These are opinions based on processes I have operated and troubleshot in the past.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

You are correct in your assessment so far. It is important to make sure that the outer loop is 3 to 10 times slower than the inner loop. The actual value varies from engineer to engineer and with situation. It is possible to have the outer loop moving faster than the inner loop. This typically leads to very poor control action. Unfortunately, I've also see that happen in management a few times with similar results on the organization.
As for competing loops, they should be reconstructed to have the competition resolved at a higher level. Each control can optimize for one variable, however complex. Optimizing for two would require a solution which eventually gets down to a single variable.
Michael
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.