A CEGAR approach for the reachability analysis of PLC-controlled chemical plants

Abstract

In this paper we address the safety analysis of chemical plants controlled by programmable logic controllers (PLCs). We consider sequential function charts (SFCs)for the programming of the PLCs, extended with the specification of the dynamic plant behavior. The resulting hybrid SFC models can be transformed to hybrid automata, opening the way to the application of advanced techniques for their reachability analysis. However, the hybrid automata models are often too large to be analyzed. To keep the size of the models moderate, we propose a counterexample-guided abstraction refinement (CEGAR) approach, which starts with the purely discrete SFC model of the controller and extends it with those parts of the dynamic behavior, which are relevant for proving or disproving safety.