Advanced

Methods and Tools for Co-Simulation of Dynamic Systems with the Functional Mock-up Interface

Andersson, Christian LU (2016) In Doctoral Theses in Mathematical Sciences
Abstract (Swedish)
Popular Abstract in Swedish

Utveckling av nya industriella produkter sker allt oftare med hjälp av datormodeller. Dessa modeller beskriver fysiken för respektive produkt, vilket exempelvis kan vara en bil eller ett kraftverk. Med hjälp av modellerna kan en stor del av analyserna på de framtida produkterna ske med hjälp av datorer, varför färre prototyper behöver konstrueras, vilket medför kortare utvecklingstider samt kostnadsbesparingar. En vanlig analys som behöver genomföras på en bil är exempelvis att undersöka hur bilen kommer att bete sig på varierande underlag eller vid olika manövreringar, dvs. man måste simulerar dess beteende. Utvecklingen mot att i högre grad använda datormodeller har pågått under en längre tid,... (More)
Popular Abstract in Swedish

Utveckling av nya industriella produkter sker allt oftare med hjälp av datormodeller. Dessa modeller beskriver fysiken för respektive produkt, vilket exempelvis kan vara en bil eller ett kraftverk. Med hjälp av modellerna kan en stor del av analyserna på de framtida produkterna ske med hjälp av datorer, varför färre prototyper behöver konstrueras, vilket medför kortare utvecklingstider samt kostnadsbesparingar. En vanlig analys som behöver genomföras på en bil är exempelvis att undersöka hur bilen kommer att bete sig på varierande underlag eller vid olika manövreringar, dvs. man måste simulerar dess beteende. Utvecklingen mot att i högre grad använda datormodeller har pågått under en längre tid, men har accelererat de senaste åren.



I en komplex datormodell, som av en bil, är det vanligt att olika fysikaliska domäner, t.ex. mekaniken och elektroniken, modelleras separat. Dessa separata modeller skapas vanligtvis med hjälp av olika verktyg. För att sedan kunna analysera den komplexa bilmodellen måste de separata modellerna kopplas samman, vilket leder till svårigheter eftersom modellerna behöver kunna utbytas mellan olika verktyg som representerar modellerna på olika vis.



Med hjälp av en ny standard har emellertid utbytet av modeller mellan olika verktyg blivit lättare - eller helt enkelt möjligt. Att koppla ihop dessa modeller har lett till att det behövs nya algoritmer och fördjupad kunskap i hur man på bästa sätt simulerar dem tillsammans. Risken är att resultatet man får inte går att lita på eller att algoritmen misslyckas med simuleringen.



Till detta behövs lättillgänglig mjukvara med tillgång till algoritmer som är fördelaktiga för olika typer av modeller. Det här är inte bara viktigt för forskning och industrin utan även i undervisningssyfte.



I denna avhandling behandlas simulering av sammankopplade modeller, där modellerna följer den nya standarden. Fokus i avhandlingen har varit på att analysera olika algoritmer, föreslår nya algoritmer, samt utveckla en mjukvara för simulering. Mjukvaran som har tagits fram har gjorts publik. (Less)
Abstract
Simulation of coupled dynamical systems, where the subsystems are bundled with their own internal solver, is important in industry. This is due to that in many cases, with complex systems, this is the only viable option. In this setting, the dynamics of each system is hidden and information between subsystems is exchanged through sampled inputs and outputs, a weakly coupled system. While a new industrial standard for exchanging models, the Functional Mock-up Interface (FMI), gains increasing acceptance, the numerical consequences of treating complex systems in this way are not completely understood.



In this thesis, stability questions of weakly coupled linear systems with feed-through are studied. Methods, within scope... (More)
Simulation of coupled dynamical systems, where the subsystems are bundled with their own internal solver, is important in industry. This is due to that in many cases, with complex systems, this is the only viable option. In this setting, the dynamics of each system is hidden and information between subsystems is exchanged through sampled inputs and outputs, a weakly coupled system. While a new industrial standard for exchanging models, the Functional Mock-up Interface (FMI), gains increasing acceptance, the numerical consequences of treating complex systems in this way are not completely understood.



In this thesis, stability questions of weakly coupled linear systems with feed-through are studied. Methods, within scope of the FMI, are further proposed which offer improved stability properties, compared to the classical case.



A simulation of a weakly coupled system introduces discontinuities due to input changes for the internal solvers. If the internal solver is a multistep method, these discontinuities will result in that the performance of the method is poor. In this thesis, a modification of the predictor in a multistep method is proposed achieving increased performance.



Furthermore, two software packages are presented. The package PyFMI is a high-level package for working with models following the FMI and contains methods for simulation of weakly coupled systems. The package Assimulo connects different solvers under a common interface which, together with PyFMI, provides an environment for evaluating solvers on industrial models. The packages are demonstrated by various examples ranging from simple test cases to a more extensive industrial application. Additionally, they have been used to verify the proposed methods and the proposed modification to the predictor, as well as being successfully used in a number of different applications outside the authors work. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Dr. Woodward, Carol, Lawrence Livermore National Laboratory, USA
organization
publishing date
type
Thesis
publication status
published
subject
keywords
Coupled Systems, Co-Simulation, FMI, Functional Mock-up Interface, PyFMI, Assimulo, Stability, Master Algorithm
in
Doctoral Theses in Mathematical Sciences
defense location
Lecture hall MH:A, Centre for Mathematical Sciences, Sölvegatan 18, Lund University, Faculty of Engineering
defense date
2016-05-04 10:15
ISSN
1404-0034
ISBN
978-91-7623-698-7
project
LCCC
language
English
LU publication?
yes
id
7dcb80cb-af81-4eed-b501-2733bd93e6eb (old id 8865908)
date added to LUP
2016-03-23 08:53:28
date last changed
2016-11-15 11:36:40
@misc{7dcb80cb-af81-4eed-b501-2733bd93e6eb,
  abstract     = {Simulation of coupled dynamical systems, where the subsystems are bundled with their own internal solver, is important in industry. This is due to that in many cases, with complex systems, this is the only viable option. In this setting, the dynamics of each system is hidden and information between subsystems is exchanged through sampled inputs and outputs, a weakly coupled system. While a new industrial standard for exchanging models, the Functional Mock-up Interface (FMI), gains increasing acceptance, the numerical consequences of treating complex systems in this way are not completely understood. <br/><br>
<br/><br>
In this thesis, stability questions of weakly coupled linear systems with feed-through are studied. Methods, within scope of the FMI, are further proposed which offer improved stability properties, compared to the classical case.<br/><br>
<br/><br>
A simulation of a weakly coupled system introduces discontinuities due to input changes for the internal solvers. If the internal solver is a multistep method, these discontinuities will result in that the performance of the method is poor. In this thesis, a modification of the predictor in a multistep method is proposed achieving increased performance.<br/><br>
<br/><br>
Furthermore, two software packages are presented. The package PyFMI is a high-level package for working with models following the FMI and contains methods for simulation of weakly coupled systems. The package Assimulo connects different solvers under a common interface which, together with PyFMI, provides an environment for evaluating solvers on industrial models. The packages are demonstrated by various examples ranging from simple test cases to a more extensive industrial application. Additionally, they have been used to verify the proposed methods and the proposed modification to the predictor, as well as being successfully used in a number of different applications outside the authors work.},
  author       = {Andersson, Christian},
  isbn         = {978-91-7623-698-7},
  issn         = {1404-0034},
  keyword      = {Coupled Systems,Co-Simulation,FMI,Functional Mock-up Interface,PyFMI,Assimulo,Stability,Master Algorithm},
  language     = {eng},
  series       = {Doctoral Theses in Mathematical Sciences},
  title        = {Methods and Tools for Co-Simulation of Dynamic Systems with the Functional Mock-up Interface},
  year         = {2016},
}