Re: [python-control-discuss] discrete time / multiple repositories

[Gustavo G. <gor...@mi...>]

I think I've figured it out.

within the yottalab dare function:
rcond,X,w,S,T = \
        sb02od(nstates, ninputs, A, B, Q, R, 'D');

should be

X,rcond,w,S,T = \
        sb02od(nstates, ninputs, A, B, Q, R, 'D');

(a similar change applies to the yottalab care function)


On Sat, Nov 26, 2011 at 2:23 PM, Roberto Bucher <rob...@su...> wrote:
> You're right
>
> something is not correct in the solution of "dare" with your matrices. I
> have to check if the problem is in my "dare" function or in the sb02od
> function of the avventi slycot library.
>
> I'll check ASAP
>
> Best regards
>
> Roberto
>
> On 11/26/2011 06:48 PM, Gustavo Goretkin wrote:
>
> Hi Roberto,
>
> Thanks for the files.
>
> I do think I'm getting different behavior between MATLAB's dlqr and
> Yottalab's. I'm attaching two files for the different environment
> showing my results.
>
> Gustavo
>
>
>
> On Sat, Nov 26, 2011 at 12:05 AM, Roberto Bucher
> <rob...@su...> wrote:
>
> Hi Gustavo
>
> here you are the modified files:
>
> matlab.py
> xferfcn.py
> statesp.py
>
> and the last yottalab.py file.
>
> Simply substitute them in the control-0.5a src folder and reinstall the
> control package.
>
> then copy directly yottalab.py at the right place (e.g.
> /usr/local/lib/python2.7/dist-packages/)
>
> Basically I've compared the result of my functions with Matlab and
> Scicoslab, and in my examples I didn't find differences.
>
> Best regards
>
> Roberto
>
> On 11/25/2011 09:48 PM, Gustavo Goretkin wrote:
>
> Hi Roberto,
>
> Thanks for the quick reply!
>
> Regarding the dlqr function in yottalab.py -- it doesn't seem to
> depend on the sampling time of the discrete time system and this
> appears to produce incorrect result, or perhaps maybe I am misusing
> the function.
>
> Thank you,
> Gustavo
>
> On Fri, Nov 25, 2011 at 3:38 PM, Roberto Bucher<rob...@su...>
>  wrote:
>
> Hi Gustavo
>
> yottalab.py requires some changes in some python control files. I'll
> check ASAP my last modifications and I send you the required files.
>
> Best regards
>
> Roberto
>
> On 11/25/2011 09:21 PM, Gustavo Goretkin wrote:
>
> Hey list,
>
> I'd like to try use some discrete-time in python-control.
> Specifically, I want MATLAB's c2d and dlqr.
>
> I'm aware of the discussion here [1] and code here [2]. yottalab.py
> has a all the functionality I want (thanks!) but I'm not sure which
> branch of python-control it is assuming. Specifically in c2d, line
> line 173, there is a call to construct a statespace object which takes
> 5 arguments A,B,C,D,Ts, but no such constructor seems to exist in
> python-control. From the rest of the code (i.e. d2c), I infer that all
> it seems to do is add a field called "Tsamp" to the statespace object.
>
> If the statespace object can represent both continuous and
> discrete-time systems, what then should the semantics be? It looks
> like  Tsamp=0 for continuous time systems.
>
> It also looks like there is a repository here [3]. Is it going to be
> merged with the sourceforge repository eventually, or other way
> around?
>
> Thanks!
> Gustavo
>
> [1] http://mail.scipy.org/pipermail/scipy-dev/2010-August/015468.html
> and
> [2] www.dti.supsi.ch/~bucher/wp-content/uploads/2011/03/yottalab.py
> [3] https://bitbucket.org/eike_welk/python-control
>
>
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure
> contains a definitive record of customers, application performance,
> security threats, fraudulent activity, and more. Splunk takes this
> data and makes sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure
> contains a definitive record of customers, application performance,
> security threats, fraudulent activity, and more. Splunk takes this
> data and makes sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure
> contains a definitive record of customers, application performance,
> security threats, fraudulent activity, and more. Splunk takes this
> data and makes sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure
> contains a definitive record of customers, application performance,
> security threats, fraudulent activity, and more. Splunk takes this
> data and makes sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
>
>
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure
> contains a definitive record of customers, application performance,
> security threats, fraudulent activity, and more. Splunk takes this
> data and makes sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
>
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
>
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure
> contains a definitive record of customers, application performance,
> security threats, fraudulent activity, and more. Splunk takes this
> data and makes sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-novd2d
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
>