Software Framework for State Estimation
Tavakoli, Siavash (2017)
2017
Automation Engineering
Teknisten tieteiden tiedekunta - Faculty of Engineering Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2017-04-05
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201703221210
https://urn.fi/URN:NBN:fi:tty-201703221210
Tiivistelmä
Over the past decade, robotics has seen tremendous increase in complexity and variety of applications. The key area in the robots seeing rapid evolution is the software. However, usually the software developed for robots has been limited to a specific application and/or a specific hardware. Unfortunately most of the software developed for robotic applications are not easily re-usable in another project. Very little effort has been done to tackle this issue and the software is developed on an ad-hoc basis.
In this work, a framework for developing sensor fusion software is proposed that is based on practices of model-driven engineering. A small domain-specific language is developed that effectively hides the lower level implementation details and makes the software development more structured and easier to re-use.
It is also discussed how graphical models can be used as computational framework for performing the statistical inference in filtering problems. It is shown how a simple estimation problem can be solved using graphical models.
In this work, a framework for developing sensor fusion software is proposed that is based on practices of model-driven engineering. A small domain-specific language is developed that effectively hides the lower level implementation details and makes the software development more structured and easier to re-use.
It is also discussed how graphical models can be used as computational framework for performing the statistical inference in filtering problems. It is shown how a simple estimation problem can be solved using graphical models.