Distributed diagnosis of discrete-event systems under dynamic event observations

In general, observable events are always assumed to be observed when studying the distributed diagnosis of discrete-event systems (DESs). However, this assumption may not be established in several DESs. In this paper, we address the problem of distributed fault diagnosis under dynamic event observations (DEOs) where some occurrences of observable events can not be captured, depending on the traces executed by systems. Then, the notion of distributed fault diagnosis under DEOs is formalized by introducing the notion of joint diagnosability with respect to DEOs (w.r.t. DEOs). For each local diagnoser, one algorithm for constructing communication models under DEOs is proposed. Finally, by constructing extended models above, an algorithm for testing joint diagnosability w.r.t. DEOs is presented. (c) 2023 The Franklin Institute. Published by Elsevier Inc. All rights reserved.