Genetic Algorithm for Generating Counterexample in Stochastic Model Checking

Abstract

Counterexamples are the most effective feature to convince system engineers about the value of formal verification. Generating the smallest counterexample in stochastic model checking has been proved to be NP-complete. In this paper, we apply the genetic algorithm to generate a counterexample for stochastic model checking. We use the diagnostic subgraph to represent a counterexample and employs indirect coding method to generate the more effective path. We implemented our method based on the stochastic model checker PRISM and applied it to some cases, in order to illustrate its applicability.