An Online AUV Trajectory Re-planning Software Architecture Based on the MOOS

Uloženo v:
Podrobná bibliografie
Název: An Online AUV Trajectory Re-planning Software Architecture Based on the MOOS
Autoři: Barišić, Matko, Vukić, Zoran
Informace o vydavateli: 2009.
Rok vydání: 2009
Témata: concurrency control, Microcomputer-based control, computer architectures, computer controlled systems
Popis: This paper discusses an open source navigation system architecture uniquely suited to use in autonomous underwater vehicles. It is based on the Mission Oriented Operating System developed by Newman (2006), Newman (2008). Some of its most pronounced advantages are an orientation towards providing a tool set for rapid prototyping of new control algorithms and support for development- stage software design in AUV systems. Its advantages are a direct result of the MOOS’ s completely modular nature, and the presented architecture has been built to preserve and, where possible, enhance this modularity. Using the MOOS design templates, control algorithms are encoded as applications “ living” in separate processes of the operating system kernel. This methodology provides for a level of robustness instrumental in autonomous vehicle since failures and errors will cause only the individual modules that incur them to fail and corresponding processes to be “ killed” by the operating system. Such critical occurrences will thereby be contained and their propagation halted from completely freezing even the low-level control loops and decision-making processes needed to successfully retrieve the malfunctioning AUV.
Druh dokumentu: Conference object
Přístupové číslo: edsair.dris...01492..d116ef774aac3279f38fea9f83486763
Databáze: OpenAIRE
Popis
Abstrakt:This paper discusses an open source navigation system architecture uniquely suited to use in autonomous underwater vehicles. It is based on the Mission Oriented Operating System developed by Newman (2006), Newman (2008). Some of its most pronounced advantages are an orientation towards providing a tool set for rapid prototyping of new control algorithms and support for development- stage software design in AUV systems. Its advantages are a direct result of the MOOS’ s completely modular nature, and the presented architecture has been built to preserve and, where possible, enhance this modularity. Using the MOOS design templates, control algorithms are encoded as applications “ living” in separate processes of the operating system kernel. This methodology provides for a level of robustness instrumental in autonomous vehicle since failures and errors will cause only the individual modules that incur them to fail and corresponding processes to be “ killed” by the operating system. Such critical occurrences will thereby be contained and their propagation halted from completely freezing even the low-level control loops and decision-making processes needed to successfully retrieve the malfunctioning AUV.