Section: New Results

Symmetry and stochastic systems

We have developed a framework to efficiently solve large Markov decision problems specified by high-level Petri nets [42] . Such a specification allows to decrease the size of the MDP by the analysis of the symmetries of the model. In a purely probabilistic context, we have designed two complementary methods for handling partial symmetries in stochastic high level Petri nets and studied their efficiency on several relevant case studies [41] .