KIT | KIT-Bibliothek | Impressum | Datenschutz

A Formal Approach to Prove Compatibility in Transformation Networks

Klare, Heiko; Pepin, Aurélien; Burger, Erik; Reussner, Ralf

The increasing complexity of software and cyberphysical systems is handled by dividing the description of the system under construction into different models or views, each with an appropriate abstraction for the needs of specific roles. Since all such models describe the same system, they usually share an overlap of information, which can lead to inconsistencies if overlapping information is not modified uniformly in all models. A well-researched approach to make these overlaps explicit and resolve inconsistencies are incremental, bidirectional model transformations. They specify the constraints between two metamodels and the restoration of consistency between their instances. Relating more than two metamodels can be achieved by combining bidirectional transformations to a network. However, such a network may contain cycles of transformations, whose consistency constraints can be contradictory if they are not aligned with each other and thus cannot be fulfilled at the same time. Such transformations are considered incompatible.
In this article, we provide a formal definition of consistency and compatibility of transformations and propose an inductive approach to prove compatibility of a given network of transformations. ... mehr

Open Access Logo

Volltext §
DOI: 10.5445/IR/1000121444
Veröffentlicht am 17.07.2020
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Programmstrukturen und Datenorganisation (IPD)
Publikationstyp Forschungsbericht/Preprint
Publikationsjahr 2020
Sprache Englisch
Identifikator ISSN: 2190-4782
KITopen-ID: 1000121444
Verlag KIT, Karlsruhe
Umfang 40 S.
Serie Karlsruhe Reports in Informatics ; 2020,3
Schlagwörter model consistency, model transformation, transformation networks, transformation decomposition, transformation compatibility
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page