Autonomous distributed system and its realization by multi agent nets

3Citations
Citations of this article
5Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Autonomous Distributed Systems (ADS) concept plays a central role for designing, operating, and maintaining complex systems in these ubiquitously networked society. Design objectives such as optimality, reliability, and efficiency are renovated according to this system paradigm. Petri nets and its related models provide one of the most concrete basis for realizing ADS to cope with their nondeterministic, concurrent, and asynchronous behavioral features. On the other hand, autonomous decisions by distributed units based on their own interests should be coordinated with total system objectives. Multi Agent Nets are Object-Oriented Colored Petri Nets for implementing autonomous intelligent units and collaborating actions among distributed units. Here in this paper, the realization of ADS by Multi Agent Nets are described through several industrial applications and prototyping that shows paramount versatility of the approach to hardware-software distributed systems encountered in wide variety of engineering problems. © Springer-Verlag Berlin Heidelberg 2007.

Cite

CITATION STYLE

APA

Kumagai, S., & Miyamoto, T. (2007). Autonomous distributed system and its realization by multi agent nets. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4546 LNCS, pp. 3–19). Springer Verlag. https://doi.org/10.1007/978-3-540-73094-1_2

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free