Re: Mixed process dynamics in DMC

[ Follow Ups ] [ Post Followup ] [ Control Consulting ] [ FAQ ]

Posted by Sean Goodhart on November 14, 19101 at 18:24:51:

In Reply to: Re: Mixed process dynamics in DMC posted by Dan O'Connor on November 09, 19101 at 17:52:53:

Dan you are exactly right that subspace does not differentiate the data if you use the detrending filter... but a time constant of 9999????!!

Seriously though you have hopefully clarified the issue for Adrian: ramp variables can have MVs which have steady state
effects and, as such, have zero rate of change as the 'steady state'. These are common - I know of several distillation examples.
DMCplus does not have a problem with such models, but as an engineer commissioning a controller with 'dynamic only' effects I would urge
caution to ensure that the MV does not become too aggressive. (Dynamic moves to help the ramp CV might cause difficulties in other CVs.)

Follow Ups:

Post a Followup

Name    : 
E-Mail  : 
Subject : 

Comments:
: Dan you are exactly right that subspace does not differentiate the data if you use the detrending filter... but a time constant of 9999????!!

: Seriously though you have hopefully clarified the issue for Adrian: ramp variables can have MVs which have steady state
: effects and, as such, have zero rate of change as the 'steady state'.  These are common - I know of several distillation examples.
: DMCplus does not have a problem with such models, but as an engineer commissioning a controller with 'dynamic only' effects I would urge
: caution to ensure that the MV does not become too aggressive.  (Dynamic moves to help the ramp CV might cause difficulties in other CVs.)


Optional Link URL:  
Link Title:         
Optional Image URL:

[ Follow Ups ] [ Post Followup ] [ Control Consulting ] [ FAQ ]