GSoC 2015 Applicant
16 views
Skip to first unread message
Renato Orsino
Mar 27, 2015, 9:47:19 AM3/27/15
to py...@googlegroups.com
Hello,
My name is Renato Maia Matarazzo Orsino, a PhD student from University of São Paulo, Brazil, and I am currently working as a guest in the Bicycle Lab of TU Delft (The Netherlands) under the supervision of Professor PhD Arend Schwab.
I am working on a new modular approach for modelling multibody systems (which can be used both on symbolic modelling and numerical simulations). In this approach you consider a complex multibody system as consisting of a set of constrained subsystems. Each subsystem can be either a single body or another "simpler" multibody system.
Consider that you already know the mathematical model of these subsystems (no matter which variables have been used for the modelling, not the formalism which has been used on its derivation, eg. Lagrange, Kane, etc.., of any of them).
Consider also that you are able to describe the constraints among these subsystems (which can be referred as "external constraints") using the modelling variables from the subsystems involved, then, by an algorithm based on finding the orthogonal complements of matrices which are associated with the jacobians of these external constraints, it is possible to obtain directly from the already known motion equations of the subsystems, the system of motion equations for the whole system (with as much equations as the number of degrees of freedom of the system) without the need of including or eliminating any constraint force.
I have already developed all the theoretical derivations of this methodology as well as some modelling and simulation examples using Mathematica and Octave. Now I want to have implemented on SymPy so that it can be shared with Multibody Systems community.
My fellow André Garnier Coutinho will also be applying once he has developed some specialized versions of this methodology to the modelling of serial and parallel robots.
Thank you for your attention.
Renato Maia Matarazzo Orsino
My name is Renato Maia Matarazzo Orsino, a PhD student from University of São Paulo, Brazil, and I am currently working as a guest in the Bicycle Lab of TU Delft (The Netherlands) under the supervision of Professor PhD Arend Schwab.
I am working on a new modular approach for modelling multibody systems (which can be used both on symbolic modelling and numerical simulations). In this approach you consider a complex multibody system as consisting of a set of constrained subsystems. Each subsystem can be either a single body or another "simpler" multibody system.
Consider that you already know the mathematical model of these subsystems (no matter which variables have been used for the modelling, not the formalism which has been used on its derivation, eg. Lagrange, Kane, etc.., of any of them).
Consider also that you are able to describe the constraints among these subsystems (which can be referred as "external constraints") using the modelling variables from the subsystems involved, then, by an algorithm based on finding the orthogonal complements of matrices which are associated with the jacobians of these external constraints, it is possible to obtain directly from the already known motion equations of the subsystems, the system of motion equations for the whole system (with as much equations as the number of degrees of freedom of the system) without the need of including or eliminating any constraint force.
I have already developed all the theoretical derivations of this methodology as well as some modelling and simulation examples using Mathematica and Octave. Now I want to have implemented on SymPy so that it can be shared with Multibody Systems community.
My fellow André Garnier Coutinho will also be applying once he has developed some specialized versions of this methodology to the modelling of serial and parallel robots.
Thank you for your attention.
Renato Maia Matarazzo Orsino
Jason Moore
Mar 27, 2015, 11:23:34 AM3/27/15
to py...@googlegroups.com
Renato,
That sounds like a very cool idea. We'd likely love to have it in SymPy/PyDy.I'd suggest this:
1. Get your application in to Melange as soon as possible. Have a look at other applications to see what people are doing.
2. You should apply to multiple orgs with the same application: PSF (SymPy), NumFOCUS (SymPy), PSF (PyDy). Make sure the title of you application reflects either SymPy or PyDy.
3. Pick an easy to fix bug in SymPy and submit a Pull Request for it. If you know Oliver, get him to help you if you haven't done this before. Or ask us questions here or on gitter.
Good luck!
--
You received this message because you are subscribed to the Google Groups "PyDy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydy+uns...@googlegroups.com.
To post to this group, send email to py...@googlegroups.com.
Visit this group at http://groups.google.com/group/pydy.
For more options, visit https://groups.google.com/d/optout.
Jason Moore
Mar 27, 2015, 11:34:19 AM3/27/15
to py...@googlegroups.com
Easy to fix bugs can be found here:
https://github.com/sympy/sympy/issues?q=is%3Aopen+is%3Aissue+label%3A%22Easy+to+Fix%22
And here are some especially easy ones:https://github.com/sympy/sympy/issues?q=is%3Aopen+is%3Aissue+label%3A%22Easy+to+Fix%22
Renato Orsino
Mar 27, 2015, 11:42:05 AM3/27/15
to py...@googlegroups.com
Jason,
Thank you so much for the information.
--
You received this message because you are subscribed to a topic in the Google Groups "PyDy" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pydy/F21ny08dvaA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pydy+uns...@googlegroups.com.
Jason Moore
Mar 27, 2015, 11:50:29 AM3/27/15
to py...@googlegroups.com
My recommendation is to upload something to Melange now. You can edit it there until the deadline.
Google will not accept late applications even if Melange stopped working.Reply all
Reply to author
Forward
0 new messages