Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Multi-Threaded Distributed System Simulations: Using Bi-Lateral Delay Lines
Linköping University, Department of Management and Engineering, Fluid and Mechatronic Systems. Linköping University, The Institute of Technology.
2013 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

As the speed increase of single-core processors keeps declining, it is important to adapt simulation software to take advantage of multi-core technology. There is a great need for simulating large-scale systems with good performance. This makes it possible to investigate how different parts of a system work together, without the need for expensive physical prototypes. For this to be useful, however, the simulations cannot take too long, because this would delay the design process. Some uses of simulation also put very high demands on simulation performance, such as real-time simulations, design optimization or Monte Carlo-based sensitivity analysis. Being able to quickly simulate large-scale models can save much time and money.

The power required to cool a processor is proportional to the processor speed squared. It is therefore no longer profitable to keep increasing the speed. This is commonly referred to as the "power wall". Manufacturers of processors have instead begun to focus on building multi-core processors consisting of several cores working in parallel. Adapting program code to multi-core architectures constitutes a major challenge for software developers.

Traditional simulation software uses centralized equation-system solvers, which by nature are hard to make parallel. By instead using distributed solvers, equations from different parts of the model can be solved simultaneously. For this to be effective, it is important to minimize overheadcosts and to make sure that the workload is evenly distributed over the processor cores.

Dividing an equation system into several parts and solving them separately means that time delays will be introduced between the parts. If these occur in the right locations, this can be physically correct, since it also takes some time for information to propagate in physical systems. The transmission line  element method (TLM) constitutes an effective method for separating system models by introducing impedances between components, causing physically motivated time delays.

Contributions in this thesis include parts of the development of the new generation of the Hopsan simulation tool, with support for TLM and distributed solvers. An automatic algorithm for partitioning models has been developed. A multi-threaded simulation algorithm using barrier synchronization has also been implemented.

Measurements of simulation time confirm that the simulation time is decreased almost proportionally to the number of processor cores for large models. The decrease, however, is reduced if the cores are divided on different processors. This was expected, due to the communication delay for processors communicating over shared memory. Experiments on real-time systems with four cores show that a four times as large model can be simulated without losing real-time performance.

The division into distributed solvers constitutes a sort of natural cosimulation. A future project could be to use this as a platform for linking different simulation tools together and simulating them with high performance. This would make it possible to model each part of the system in the most suitable tool, and then connect all parts into one large model.

Place, publisher, year, edition, pages
Linköping: Linköping University Electronic Press, 2013. , 56 p.
Series
Linköping Studies in Science and Technology. Thesis, ISSN 0280-7971 ; 1576
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:liu:diva-88025Local ID: LIU-TEK-LIC-2013:10ISBN: 978-91-7519-694-7 (print)OAI: oai:DiVA.org:liu-88025DiVA: diva2:601342
Presentation
2013-02-08, A34, Hus A, Campus Valla, Linköpings universitet, Linköping, 10:15 (English)
Opponent
Supervisors
Available from: 2013-01-29 Created: 2013-01-29 Last updated: 2016-10-31Bibliographically approved
List of papers
1. Next Generation Simulation Software using Transmission Line Elements
Open this publication in new window or tab >>Next Generation Simulation Software using Transmission Line Elements
Show others...
2010 (English)In: Fluid Power and Motion Control / [ed] Dr D N Johnston and Professor A R Plummer, Centre for Power Transmission and Motion Control , 2010, 265-276 p.Conference paper, Published paper (Refereed)
Abstract [en]

A suitable method for simulating large complex dynamic systems is represented by distributed modelling using transmission line elements. The method is applicable to all physical systems, such as mechanical, electrical and pneumatics, but is particularly well suited to simulate systems where wave propagation is an important issue, for instance hydraulic systems. By using this method, components can be numerically isolated from each other, which provide highly robust numerical properties. It also enables the use of multi-core architecture since a system model can be composed by distributed simulations of subsystems on different processor cores.

Technologies based on transmission lines has successfully been implemented in the HOPSAN simulation package, develop at Linköping University. Currently, the next generation of HOPSAN is developed using an object-oriented approach. The work is focused on compatibility, execution speed and real-time simulation in order to facilitate hardware-in-the-loop applications. This paper presents the work progress and some possible features in the new version of the HOPSAN simulation package.

Place, publisher, year, edition, pages
Centre for Power Transmission and Motion Control, 2010
National Category
Mechanical Engineering
Identifiers
urn:nbn:se:liu:diva-59661 (URN)978-1-86197-181-4 (ISBN)
Conference
Fluid Power and Motion Control, 15th-17th September, Bath, England, UK
Projects
HiPO
Available from: 2010-11-08 Created: 2010-09-23 Last updated: 2016-05-27Bibliographically approved
2. High Performance System Simulation Using Multiple Processor Cores
Open this publication in new window or tab >>High Performance System Simulation Using Multiple Processor Cores
2011 (English)In: The Twelfth Scandinavian International Conference on Fluid Power, SICFP'11 / [ed] Harri Sairiala & Kari T. Koskinen, 2011Conference paper, Published paper (Refereed)
Abstract [en]

Future research and development will depend on high-speed simulations, especially for large and complex systems. Rapid prototyping, optimization and real-time simulations require  simulation tools that can take full advantage of  computer hardware.  Recent developments  in the computer market indicate  a change in focus from increasing the speed of processor cores towards increasing the number of cores in each processor. HOPSAN is a simulation tool for fluid power and mechatronics, developed at Linköping University. It  is based upon the transmission line  modeling  (TLM)  technique. This method is very suitable for taking advantage of multi-core  processors.  This paper presents  the  implementation  of multi-core support in the next generation of HOPSAN. The concept is to divide the  model  into equally sized  groups of  independent components,  to make it possible to  simulate  them  in separate threads. Reducing overhead costs and finding an effective sorting algorithm constitute  critical steps for maximizing the benefits.  Experimental results show  a significant reduction in simulation time. Improvement of algorithms in combination with a continuous increase in the number of processor cores can potentially  lead to further  increases  in simulation performance. 

Keyword
Multi-core, simulation, transmission line element method, transmission line modelling, fluid power, system simulation
National Category
Engineering and Technology
Identifiers
urn:nbn:se:liu:diva-68373 (URN)978-952-15-2517-9 (ISBN)978-952-15-2520-9 (ISBN)978-952-15-3273-3 (ISBN)
Conference
The Twelfth Scandinavian International Conference on Fluid Power, SICFP'11, 18th–20th May, Tampere, Finland
Available from: 2011-05-23 Created: 2011-05-23 Last updated: 2016-04-07
3. Multi-Threaded Real-Time Simulations of Fluid Power Systems Using Transmission Line Elements
Open this publication in new window or tab >>Multi-Threaded Real-Time Simulations of Fluid Power Systems Using Transmission Line Elements
2012 (English)Conference paper, Published paper (Refereed)
Abstract [en]

The demand for large-scale real-time simulations of fluid power systems is in-creasing, due to growing demands for added functionality. Real-time simulationscan be used in for example hardware-in-the-loop experiments and embeddedcontrol systems. In order to achieve real-time performance, it is often necessaryto use small or simplified models, reducing the usefulness and accuracy of theresults. This article proposes the use of transmission line modelling (TLM) forexploiting multi-core hardware in real-time and embedded systems. The charac-teristics of the TLM method are analysed to identify difficulties and possibilities.A method for how to parallelise TLM models is then presented. Subsequently, aprogramming interface for implementing the parallel models in the target systemsis introduced. Practical experiments show that the approach works and that themethod is applicable. So far, however, it has required great effort on the part ofthe engineer, both when it comes to programming, compiling and importing themodel into the target environments, although some attempts to automate the pro-cedure have been successful, reducing the level of complexity.

Keyword
Real-time simulation, Distributed modelling, Transmission line mod- elling, Parallel simulation, Multi-core, Model fidelity
National Category
Fluid Mechanics and Acoustics Other Mechanical Engineering
Identifiers
urn:nbn:se:liu:diva-76377 (URN)
Conference
8th International Fluid Power Conference, March 26-28, 2012, Dresden, Germany
Available from: 2012-04-05 Created: 2012-04-05 Last updated: 2015-11-19Bibliographically approved
4. Multi-Threaded Distributed System Simulations Using the Transmission Line Element Method
Open this publication in new window or tab >>Multi-Threaded Distributed System Simulations Using the Transmission Line Element Method
2016 (English)In: Simulation (San Diego, Calif.), ISSN 0037-5497, E-ISSN 1741-3133, Vol. 92, no 10, 921-930 p.Article in journal (Other academic) Published
Abstract [en]

By introducing physically motivated time delays, simulation models can be partitioned into decoupled independent sub-models. This enables parallel simulations on multi-core processors. An automatic algorithm is used for partitioning and running distributed system simulations. Methods for sorting and distributing components for good load balancing have been developed. Mathematical correctness during simulation is maintained by a busy-waiting thread synchronization algorithm. Independence between sub-models is achieved by using the transmission line element method. In contrast to the more commonly used centralized solvers, this method uses distributed solvers with physically motivated time delays, making simulations inherently parallel. Results show that simulation speed increases almost proportionally to the number of processor cores in the case of large models. However, overhead time costs mean that models need to be over a certain size to benefit from parallelization.

Place, publisher, year, edition, pages
Sage Publications, 2016
Keyword
Distributed Solvers, Parallelism, Problem Partitioning, Transmission Line Modelling, System Simulation
National Category
Engineering and Technology
Identifiers
urn:nbn:se:liu:diva-88024 (URN)10.1177/0037549716667243 (DOI)000385704300004 ()
Note

When first pubished online the status of this article was Manuscript.

Funding agencies: ProViking research School; Swedish Foundation for Strategic Research (SSF)

Available from: 2013-01-29 Created: 2013-01-29 Last updated: 2017-12-06Bibliographically approved

Open Access in DiVA

Multi-Threaded Distributed System Simulations: Using Bi-Lateral Delay Lines(1201 kB)1242 downloads
File information
File name FULLTEXT01.pdfFile size 1201 kBChecksum SHA-512
14dcff4e9bad764aecf3137c3305ab7f663ef3638140fd1e277695a04dc34de0fdc8f2ec6c64dbfad2d27e5eb6b31fef48e4f32f4a38622f4b5682bb6cb6a46d
Type fulltextMimetype application/pdf
omslag(206 kB)43 downloads
File information
File name COVER01.pdfFile size 206 kBChecksum SHA-512
d95e076b2608465684e48bf89cd49fdd135f97ee523be0b923d26e78c4411502edc3191ad5aabc71a6144e7d988173775f0ebbba6bf6854ff9adb8d4908c56ad
Type coverMimetype application/pdf

Search in DiVA

By author/editor
Braun, Robert
By organisation
Fluid and Mechatronic SystemsThe Institute of Technology
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 1242 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 1525 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf