Rule-based Composable Modelling of Industrial Automation Automata under Nominal and Faulty Conditions

Abstract

In this paper a novel modelling methodology for industrial automation discrete event dynamical systems is presented and some preliminary results are drawn by means of a significative example. The proposed approach exploits behavioral rules expressed by means of regular languages generated by automata, and composed using automata parallel composition. In this framework, with a suitable rules definition, fault events will cause a changing in a limited set of rules (i.e automata), hence a sort of “fault modelling confinement” can be achieved. This approach leads to a sort of “multi-model” set of parallel automata representing a system in nominal and different faulty conditions, hence statecharts have been effectively used for representation. The resulting models are suitable for formal analysis and design based on well-established tools for discrete event dynamical systems. In addition, the “complexity explosion”, which is rather common for the models adopted in rigorous treatment of discrete event systems, can be effectively avoided exploiting composability properties and simple and “confined” fault modelling.