Modelling of UML sequence diagrams with generalized nets

Koycheva, Evelina; Trifonov, Trifon; Aladjov, Hristo

doi:10.1109/is.2002.1042592

Cited by 4 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generalized Nets [GN,9] are proven to meet all requirements above [10,11] -e.g. resources exist there as transitions, behavior as places and their characteristic functions, and system users as tokens.…”

Section: The Mof-based Meta-model Of Gnsmentioning

confidence: 99%

Integrating analysis capabilities into the model driven engineering process

Koycheva

Hennig

Braune

2010

2010 5th IEEE International Conference Intelligent Systems

View full text Add to dashboard Cite

The state-of-the-art object-oriented modeling language UML does not offer any possibilities for model based analysis of functional or non-functional properties of a future solution. Reasons therefore are the semi-formality of its specification, the fact that it is not an executable language (yet) as well as the existence of informational gaps in the data, required for the desired analysis. This disadvantage can be balanced out without losing a certain level of conformity at modeling time if the available models can be transformed into another domain, which is suitable for analysis purposes. Since UML is defined in a MOF manner a high level of conformance to the target domain can be achieved by defining the target model MOF-conformal, too. Thereby a high level of automation and universality of the transformation rules can be reached if the domains are specified on the highest abstract level as possible, i.e. on MOF architectural level M2, which represents their meta-models. For obtaining unambiguous results by the analysis the target domain of the transformation has to be a formal one. Generalized ets as proven means for evaluation meet this requirement. Hence this paper represents the meta-model of the Generalized ets in MOF manner. Thus models, based on UML and its lightweight extensions like the MARTE profile, can be converted in an analyzable domain. This approach, focusing on an evaluation and optimization in the early design phases, promises the considerable reduction of project costs and the prevention of time overruns.

show abstract

Section: The Mof-based Meta-model Of Gnsmentioning

confidence: 99%