Distributed synchronous diagnosis of discrete event systems modeled as automata. (October 2021)

• Record Type: Journal Article
• Authors: Veras, Maria Z.M.; Cabral, Felipe G.; Moreira, Marcos V.
• Is Part Of: Control engineering practice. Volume 115(2021)
• Publication Date: 2021-10
• Subjects: Distributed diagnosis -- Discrete-event systems -- Automata; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2021.104892
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: Recently, the decentralized synchronous diagnosis (DESD) method has been proposed for Discrete-Event Systems composed of several modules or subsystems, where the size of the local diagnosers grows linearly with the size of the fault-free behavior models of the system components. Thus, the memory space required to implement the local diagnosers in a computer is reduced in comparison with traditional diagnosis strategies that are based on the composed system model, which may grow exponentially with the number of system modules. The main drawback of the DESD strategy is the possibility of acceptance of an exceeding language as part of the fault-free system behavior by the diagnosis scheme. This exceeding language can be associated with non-detectable faults or an increase in the diagnosis delay bound. In this paper, we propose a distributed synchronous diagnosis (DISD) architecture, where the local diagnosers are implemented considering a specific communication protocol that refines the state estimate of the fault-free behavior of the system modules, leading to a smaller exceeding language accepted as fault-free by the DISD scheme in comparison with the DESD scheme. As in the DESD, in the DISD, the local diagnosers do not grow exponentially with the number of system modules, requiring smaller memory space for implementation than traditional methods. We also define the property of distributed synchronous diagnosability, and propose a verification method. Highlights: Distributed diagnosis scheme for Discrete-Event Systems composed of several modules. The memory space needed to store the local diagnosers is smaller than other methods. The local diagnosers communicate with other local diagnosers in the same network. The communication between diagnosers can be used to improve fault diagnosis.