Re: [Orekit Developers] Kalman Filter

[MAISONOBE Luc <luc.maisonobe@c-s.fr>]

Pau Hebrero Casasayas <Pau.Hebrero-Casasayas@scisys.de> a écrit :

> Dear Luc, Paul,
>
> Thanks a lot for your help on the Kalman Filter topic. I will come  
> back whenever any other related questions arise.

Please wait a few days before looking at this. The refactoring is
still ongoing between Orekit and Hipparchus. Hopefully, it will be
ready very soon now. I'll notify the list when ready for review.

best regards,
Luc

> Again, thanks a lot,
>
> Pau
>
> ----------------------------------------------------------------------------------------
> Pau Hebrero Casasayas
> Flight Dynamics Software Engineer
> Space
> SCISYS Deutschland GmbH
> T: +49 6151 4285343 | F: +49 6151 4285399
> E: Pau.Hebrero-Casasayas@scisys.de | http://www.space.scisys.de
>
> SCISYS Deutschland GmbH, Darmstadt Office, Europaplatz 4, 64293  
> Darmstadt, Germany
> Management Board: Prof. Dr. Klaus-G. Meng (Chairman), Sandra  
> Krewerth, Ulli Leibnitz, Dr. Karl-W. Pieper, Dr. Horst Wulf
> Registered Office Bochum HRB 13694, VAT-IdNo DE 813242674,  
> WEEE-Reg.-No. DE 74530735
> -----Original Message-----
> From: orekit-developers-request@orekit.org  
> [mailto:orekit-developers-request@orekit.org] On Behalf Of MAISONOBE  
> Luc
> Sent: 13 March 2018 20:16
> To: orekit-developers@orekit.org
> Subject: Re: [Orekit Developers] Kalman Filter
>
>
> Pau Hebrero Casasayas <Pau.Hebrero-Casasayas@scisys.de> a écrit :
>
> > Hi all,
>
> Hi Pau,
>
> > I have read that there is a Kalman Filter currently under development.
> > We would like to do some real-time orbit determination testing using a
> > Kalman Filter, mainly we are after the noise generated after the OD.
> >
> > Is there a branch that I can check-out to try it? To which extend is
> > it reliable?
>
> Yes, the branch in the main git repository is called kalman-filter-od.
> It can be seen online here :
> <https://www.orekit.org/forge/projects/orekit/repository?utf8=%E2%9C%93&rev=kalman-filter-od>, but simply cheking it out from git is  
> simpler.
>
> It is reliable but the API will change very soon. We have started to  
> refactor the mathematical part so it can be contributed back to  
> Hipparchus, and Orekit will simply add the space flight dynamics  
> layer on top of it. You can look at  
> <https://github.com/Hipparchus-Math/hipparchus/issues/32>
> and at the kalman branch in Hipparchus  
> <https://github.com/Hipparchus-Math/hipparchus/tree/kalman>.
> I still have to:
>    1) add the non-linear (extended) Kalman filter in Hipparchus
>    2) split the Orekit implementation so it delegates to Hipparchus
>
> The API changes in Orekit will most certainly be to introduce a  
> Streams, which are very well suited for real-time long term runs.  
> Basically, you will have a Stream of measurements as input, and you  
> will produce a Stream of Stream of Spacecraftstate as output.
>
> I think the Hipparchus stuff should be ready next week, and the  
> changes to Orekit will be started after that. We can expect to have  
> the Orekit kalman-filter-od branch to be merged back to the  
> development branch in the next few weeks.
>
> best regards,
> Luc
>
> > Many thanks in advance,
> >
> > Pau
> >
> > ___________________________________________________________
> >
> > Pau Hebrero Casasayas
> > Flight Dynamics Software Engineer
> > Space
> > SCISYS Deutschland GmbH
> > T:  +49 6151 4285343 | F:  +49 6151 4285399
> > E:
> > Pau.Hebrero-Casasayas@scisys.de<mailto:Pau.Hebrero-Casasayas@scisys.de
> > > | http://www.space.scisys.de
> >
> > SCISYS Deutschland GmbH, Darmstadt Office, Europaplatz 4, 64293
> > Darmstadt, Germany Management Board: Prof. Dr.-Ing. Klaus-G. Meng
> > (Chairman), Sandra Krewerth, Ulli Leibnitz, Dr. Karl-W. Pieper, Dr.
> > Horst Wulf Registered Office Bochum HRB 13694, VAT-Id-No. DE
> > 813242674, WEEE-Reg.-No. DE 74530735