[HN Gopher] All about Modelica: An equation-based language for m...
___________________________________________________________________
All about Modelica: An equation-based language for modeling
physical systems
Author : macro-b
Score : 92 points
Date : 2020-06-30 15:05 UTC (7 hours ago)
(HTM) web link (marcobonvini.com)
(TXT) w3m dump (marcobonvini.com)
| pflanze wrote:
| There's a project to create a language similar to Modelica as a
| DSL in Julia, Modia[1]. I'm not currently a user of either
| language, but saw a presentation at JuliaCon about it[2] and
| thought it looked interesting.
|
| [1] https://github.com/ModiaSim/Modia.jl [2]
| https://youtu.be/hVg1eL1Qkws
| zokier wrote:
| okay, its neat and all, but I missed seeing any _output_ from the
| models, or examples how you use the models you built to analyze
| something? How do you interact with it, what sort visualizations
| you get etc etc??
| ptrott2017 wrote:
| You can plot variable outputs in 2D and 3D plots, use values to
| drive simulation based animations (2D and 3D) and via
| Functional Mockup interface standard (fmi-standard.org) you can
| interconnect external simulations and interactive
| visualizations (2D, 3D, VR etc). For examples of these in
| OpenModelica see the user guide here:
| https://www.openmodelica.org/doc/OpenModelicaUsersGuide/Open...
|
| There is also a notebook UI that lets you do a lot of
| Ipython/Jupyter style dev with feedback plots and
| visualizations, in addition to code tools and visual authoring.
| amatic wrote:
| Is there a graphical interface, like Simulink, for Modelica?
| vsskanth wrote:
| OpenModelica is free. There are many commercial IDEs like
| Dymola from Dassault, Wolfram System Modeler etc.
| choeger wrote:
| Many. Dymola is probably the go-to solution, but also
| openmodelica, simulationX and wolfram, I think.
|
| Here is a list: https://www.modelica.org/tools
| ptrott2017 wrote:
| There are commercial ones like ones like Altair SolidThinking,
| Wolfram System Modeller, ESI's SimulationX and there are open
| source ones such as OMedit. Which meets your needs will be
| partly budget / workflow / toolbox support based. If your new
| to Modelica - then OpenModelica and OMEdit (OMEdit comes as
| part of the OpenModelica download) are great starting points.
| To get an idea of the available commercial and open source
| options see:
|
| https://www.modelica.org/tools
|
| and https://openmodelica.org/openmodelicaworld/tools
| ptrott2017 wrote:
| This is a great overview of Modelica. For those looking to get
| started - check out OpenModelica (https://openmodelica.org/) and
| if you are curious how to approach deploying and using Modelica
| in web services the article's author as a rather awesome open
| source project called LambdaSim
| see(https://github.com/mbonvini/LambdaSim)
| jwilber wrote:
| Stellar example of naming gone right in software.
| vsskanth wrote:
| a quick intro for programmers - Modelica is basically the
| equivalent of Haskell for engineers - the compiler does most of
| the work for you and produces highly performant code. You can
| compose a very complex system from individually modeled elements.
| The Modelica compiler will check constraints and basically come
| up with a system of equations. If it compiles, it is probably
| correct. The model wont compile if a particular variable is being
| modified in two places. Models have full introspection where you
| can observe any variable you choose.
| dang wrote:
| Is Modelica used for economic modeling as well?
| dongping wrote:
| I don't know enough about economic modeling, but if the models
| can somehow be expressed as driven by a "potential" and a
| "flow" variable (like electric voltage and electric currents),
| then it would definitely make sense to try that out in
| Modelica.
| vsskanth wrote:
| Been working in Modelica for a couple of years. Modelica
| compilers generate C code since they're heavily used in the
| embedded and real-time space. Kinda amazing to see systems with
| hundreds of thousands of equations simulate faster than real
| time.
| choeger wrote:
| One thing about Modelica is quite important and might confuse
| newcomers: the language specification mostly only covers the
| model, not the equations. That is, when you evaluate your model
| under, e.g., Dymola and Openmodelica you will (most of the time)
| end up with equal (or rather similar) systems of equations. The
| meaning of these equations, i.e., the actual simulation can
| easily differ, though (except for the most trivial models).
|
| The reason for this is that the equations are extremely powerful
| (you can express pretty much anything computable inside an
| equation) and their numerical solution is often driven by
| heuristics (e.g., finding initial values). This makes it
| extremely difficult to come up with a common standard for the
| solution of Midelica's hybrid DAEs. Hence you will find little
| written information about what exactly happens during a
| simulation that applies to more than one implementation.
| dang wrote:
| If curious see also (from 2017):
| https://news.ycombinator.com/item?id=16013179
| budadre75 wrote:
| does anyone know any industry usage of Modelica?
| vsskanth wrote:
| energy companies (powerplants, building systems etc.),
| aerospace, automotive, motorsports, robotics
| JPLeRouzic wrote:
| Maybe also in physiology (BTW I would be happy if someone
| could point out similar or better resources)
|
| https://www.physiolibrary.org/doc/Physiolibrary.html#Physiol.
| ..
| rhodin wrote:
| Check the vendors user stories: https://www.wolfram.com/system-
| modeler/customer-stories/ Check the Modelica conferences (https
| ://modelica.org/events/modelica2019/proceedings/html/Mo...) and
| you'll see a lot of automotive companies (Toyota, Volvo, Bosch,
| etc)
___________________________________________________________________
(page generated 2020-06-30 23:00 UTC)