Skip to Main content Skip to Navigation

Modular Specification and Compositional Analysis of Stochastic Systems

Benoît Delahaye 1
1 S4 - System synthesis and supervision, scenarios
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : This thesis presents new contributions in the design and verification of systems mixing both non-deterministic and stochastic aspects. Our results can be divided into three main contributions. First, we generalize interface theories to the stochastic setting. We build upon the known formalism of Interval Markov Chains to develop Constraint Markov Chains, the first fully compositional specification theory for stochastic systems. Second, we extend the notion of assume-guarantee contracts and develop a contract-based theory for stochastic systems, proposing quantitative notions for satisfaction and refinement. Finally, we propose a methodology for the verification of complex systems. This methodology is based on a stochastic abstraction of the environment where two components are working, allowing to verify the components individually. Combined with statistical model checking, this methodology is successfully applied to the verification of an industrial case study.
Document type :
Complete list of metadata

Cited literature [129 references]  Display  Hide  Download
Contributor : Benoît Delahaye <>
Submitted on : Monday, May 9, 2011 - 4:14:00 PM
Last modification on : Tuesday, June 15, 2021 - 4:27:46 PM
Long-term archiving on: : Friday, November 9, 2012 - 10:57:15 AM



  • HAL Id : tel-00591609, version 1


Benoît Delahaye. Modular Specification and Compositional Analysis of Stochastic Systems. Modeling and Simulation. Université Rennes 1, 2010. English. ⟨tel-00591609⟩



Record views


Files downloads