A Event Occurrence Rules based Compact Modeling Formalism for a Class of Discrete Event Systems (Q3147797)

In two recent papers, the authors (2001) developed a technique for obtaining automata models of discrete event systems using a rule-based modelling formalism. In that approach, a system consists of binary valued input and output (I/O) signals, the state of the system is determined solely by the I/O values, and the model complexity is polynomial in the number of signals. The model is described in terms of initial conditions, weighted sensor event occurrence rules, and precedence relations. The present paper continues the task of improving and simplifying the authors' above-mentioned work and succeeds in eliminating the precedence relations among output events by capturing them as part of the event occurrence rules. In the proposed formalism, the representation of the system is also polynomial in the size of signals, in contrast to the corresponding equivalent automata models whose state space is exponential in the size of signals. The obtained compactness of the proposed rule-based model, together with its intuitive nature, makes it user-friendly, less error-prone, more flexible, easily scalable, and provides canonicity of the representation. The flexibility of the considered modelling formalism involves rapid construction, reconfiguration, debugging, and validation of the model, and is illustrated on examples drawn from manufacturing and process control systems.