This paper shows how to abstract networks of timed automata in order to accelerate the analysis of quantitative properties such as path or cycle duration, that would otherwise suffer from the state space explosion. Two approaches are introduced, a single step strategy and an iterative one, where a part of the network of timed automata is merged and abstracted. As a consequence, the state space is reduced and model-checking is simplified. These approaches are illustrated on a case study, where the comparison is done by calculating the cycle time of one automaton in the network, both on the real network and on the two abstracted ones, showing that the method reduces significantly the runtime, or simply renders feasible the analysis of the system.
CITATION STYLE
Devillers, R., & Klaudel, H. (2016). Abstraction strategies for computing travelling or looping durations in networks of timed automata. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9884 LNCS, pp. 140–156). Springer Verlag. https://doi.org/10.1007/978-3-319-44878-7_9
Mendeley helps you to discover research relevant for your work.