Re: Convolution using intermediate variables

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

Posted by Mike McCarty on February 27, 2000 at 21:24:14:

In Reply to: Convolution using intermediate variables posted by Sri Sampath on February 15, 2000 at 21:18:59:

A "convolution" technique would normally be used when developing a response relationship of A to C, by finding a response of A to B and then B to C. You could combine (or convolute) the two intermediate responses to get a combined A to C response.

There is no method for developing a relationship of A to C by using an intermediate variable B when the variable C can not be observed by plant testing. Likewise, if the measurement for C is available, but you can not collect data in the "plant test mode" then convolution is not the proper technique. In both of these cases, one technique for developing an unobservable / uncollectable response variable is by manually building a response using engineering judgement (also called guessing). However, a much better way is to only use variables that can actually be collected, and to not use data collected during "severe disturbances".

Mike

Follow Ups:

Post a Followup

Name    :

E-Mail  :

Subject :

Comments:

: A "convolution" technique would normally be used when developing a response relationship of A to C, by finding a response of A to B and then B to C.  You could combine (or convolute) the two intermediate responses to get a combined A to C response.

: There is no method for developing a relationship of A to C by using an intermediate variable B when the variable C can not be observed by plant testing.  Likewise, if the measurement for C is available, but you can not collect data in the "plant test mode" then convolution is not the proper technique.  In both of these cases, one technique for developing an unobservable / uncollectable response variable is by manually building a response using engineering judgement (also called guessing).  However, a much better way is to only use variables that can actually be collected, and to not use data collected during "severe disturbances".

: Mike

Optional Link URL:

Link Title:

Optional Image URL:

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