bugGNU Octave - Bugs: bug #45336, [octave forge] (control) version...

 
 

bug #45336: [octave forge] (control) version 2.8.2 - error on MIMO tf multiply

Submitted by:  Nicholas Jankowski <nrjank>
Submitted on:  Wed 17 Jun 2015 02:38:23 AM UTC  
 
Category: Octave Forge PackageSeverity: 3 - Normal
Priority: 5 - NormalItem Group: Incorrect Result
Status: NoneAssigned to: None
Originator Name: Open/Closed: Open
Release: 4.0.0Operating System: Microsoft Windows

Add a New Comment(Rich Markup)
   

You are not logged in

Please log in, so followups can be emailed to you.

 

Mon 22 Jun 2015 11:18:01 AM UTC, comment #5:

The mpower operator for LTI models computes a1^n as n-1 multiplications (mtimes operator). For MIMO transfer functions, mtimes converts both factors to state-space, computes the multiplication and then converts the product back to transfer function. Therefore, mpower computes a3 = a1 * (a1 * a1).

Previously, the conversion of the transfer function to state-space and back took place during each multiplication:

With the patch, the tf -> ss conversion is the first step and the ss -> tf conversion the last step.

Obviously, the patch doesn't help if you write a1a1a1 instead of a1**3 because no tf -> ss -> tf conversion is saved.

Lukas Reichlin <paramaniac>
Sat 20 Jun 2015 07:54:46 PM UTC, comment #4:

just grabbed the revised mpower.m file. the code below in the original error report works. I noticed all the warnings about SS conversion disappeared. a1**n seemed well behaved out through n=200

I did try a1a1, a1a1*a1... since only mpower changed, that still gave the same warnings and odd results at n=3, error at n=4. so it doesn't seem restricted to mpower

Nicholas Jankowski <nrjank>
Sat 20 Jun 2015 06:36:30 PM UTC, comment #3:

Thanks for reporting, I've just pushed a quick fix:

https://bitbucket.org/paramaniac/control/commits/520c14850013ec0c7c3925d83db9c2c9fb4d8636

Lukas Reichlin <paramaniac>
Sat 20 Jun 2015 06:39:12 AM UTC, comment #2:

in addition on 2.8.2 there are 2 warnings in the test area

Testing: /usr/share/octave/packages/control-2.8.2/@iddata/ifft.m
warning: iddata: more outputs than samples - matrice 'y' should probably be tran
sposed
warning: iddata: more inputs than samples - matrice 'u' should probably be trans
posed

marco atzeri <matzeri>
Fri 19 Jun 2015 01:45:35 PM UTC, comment #1:

Ccing the control package maintainer.

Mike Miller <mtmiller>
Project Administrator
Wed 17 Jun 2015 02:38:23 AM UTC, original submission:

after upgrading to 2.8.2, a script that worked under 2.8.1 started throwing an error. here's a simpler version that produces the same error:

don't know if this is related to Bug #45334, which relates to odd outputs from multiple tf multiplies. But this error above is new to 2.8.2, and the case described in #45334 does not create this error in 2.8.1 or 2.8.2.

Nicholas Jankowski <nrjank>

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach File(s):
   
   
Comment:
   

No files currently attached

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by jwe (Updated the item)
  • -unavailable- added by paramaniac (Posted a comment)
  • -unavailable- added by matzeri (Posted a comment)
  • -unavailable- added by mtmiller (Posted a comment)
  • -unavailable- added by mtmiller
  • -unavailable- added by nrjank (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only project members can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 2 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Sun 13 Aug 2017 01:43:02 PM UTCjweSummarycontrol 2.8.2 - error on MIMO tf multiply =>[octave forge] (control) version 2.8.2 - error on MIMO tf multiply
    Fri 19 Jun 2015 01:45:35 PM UTCmtmillerCarbon-Copy-=>Added paramaniac

    Back to the top


    Powered by Savane 3.1-cleanup1