Verification of generalized robust diagnosability of discrete event systems

Abstract

This work addresses the problem of generalized robust diagnosability (GRD) of discrete event systems (DESs) described by a class of automata, where each automaton in the class generates a distinct language. The definition of GRD and the algorithm for its verification previously proposed in literature were updated, resulting in an algorithm with smaller computational complexity than the previous one. Based on this algorithm, a new necessary and sufficient condition for generalized robust diagnosability was presented. Four different approaches on diagnosability of DESs were analyzed: the problem of diagnosability of discrete event systems subject to permanent sensor failures (i); the problem of robust diagnosis of discrete event systems against permanent (ii) and intermittent (iii) loss of observations; and the problem of verification of robust diagnosability for partially observed discrete event systems (iv). Transformation mechanisms for each analyzed problem were proposed with the purpose of demonstrating that all approaches (i) - (iv) are particular cases of the generalized robust diagnosability definition proposed in this work.