Semantics and implementation of a refinement relation for behavioural models
Résumé
In this paper, we present a way to implement refinement relations over transition systems, useful for incremental behavioural model development. Our work is based on the extension relation defined over Labelled Transition Systems (LTS). This relation appears to be suitable for refinement developments, but its computability has not yet been established. We propose an implementation which relies on the generalisation of a bisimulation relation applied on acceptance graphs. It is formally demonstrated and illustrated through a case study of a system modelled both in UML state machine and LTS. The analysis is performed on LTS by a Java prototype we have developed, and the interpretation of results is given on the corresponding UML state machine. Moreover, when the relation is not satisfied, the tool can exhibit failing traces and can give several proposals to improve the model. Hence, this work goes toward a behavioural semantics for the class specialisation relation of object oriented models.
Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...